diff --git a/.github/workflows/malecns-subgraph.yml b/.github/workflows/malecns-subgraph.yml index fa6faa4..304fdb7 100644 --- a/.github/workflows/malecns-subgraph.yml +++ b/.github/workflows/malecns-subgraph.yml @@ -81,7 +81,7 @@ jobs: # small enough to belong in git; the rest are artifacts. # ------------------------------------------------------------------ - name: Build the scale ladder - if: ${{ github.event.inputs.ladder_sizes != '' }} + if: ${{ github.event_name == 'schedule' || github.event.inputs.ladder_sizes != '' }} env: SIZES: ${{ github.event.inputs.ladder_sizes || '1000,5000,10000,25000,50000' }} MINW: ${{ github.event.inputs.min_weight || '5' }} @@ -125,7 +125,7 @@ jobs: name: malecns-neighborhoods path: data/derived/malecns_*_neighborhood.json - name: Upload the scale ladder (every rung, including the large ones) - if: ${{ github.event.inputs.ladder_sizes != '' }} + if: ${{ github.event_name == 'schedule' || github.event.inputs.ladder_sizes != '' }} uses: actions/upload-artifact@v4 with: name: malecns-scale-ladder diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml index 274cdce..60a0bab 100644 --- a/.github/workflows/release.yml +++ b/.github/workflows/release.yml @@ -43,7 +43,8 @@ jobs: - name: Install run: | python -m pip install --upgrade pip wheel setuptools - python -m pip install -e ".[dev,viz]" + python -m pip install -e ".[dev,viz,browser-test]" + python -m playwright install --with-deps chromium - name: Record the environment run: | diff --git a/.github/workflows/tests.yml b/.github/workflows/tests.yml index 1de3b21..795ff32 100644 --- a/.github/workflows/tests.yml +++ b/.github/workflows/tests.yml @@ -59,7 +59,9 @@ jobs: python-version: "3.12" - run: | python -m pip install --upgrade pip wheel setuptools - python -m pip install -e ".[dev,viz]" + python -m pip install -e ".[dev,viz,browser-test]" + - name: install Chromium for the static-browser smoke check + run: python -m playwright install --with-deps chromium - name: pytest (slow suite) run: pytest -q -m slow - name: verify the artifact manifest diff --git a/CITATION.cff b/CITATION.cff index 8fbb437..fffdb7f 100644 --- a/CITATION.cff +++ b/CITATION.cff @@ -22,7 +22,8 @@ abstract: >- are demonstrated on cuts of the public MaleCNS v1.0 connectome, where the headline result is methodological: a dependence verdict obtained on one extract of a network can invert on another extract of the same network, and - the extract's recurrence rather than its size predicts which way. + the saved extracts associate the verdict change more with recurrence than node + count, without isolating recurrence as a cause. Concretely, FlyLab maps a compound and a free concentration onto a typed engagement panel for insect receptors and their vertebrate counterparts at the same dose, converts the insect side through a fixed mechanism table into gain diff --git a/HANDOFF.md b/HANDOFF.md index 2d5556f..92de213 100644 --- a/HANDOFF.md +++ b/HANDOFF.md @@ -1,8 +1,8 @@ -# FlyLab handoff (v0.6.1, 2026-09-20) +# FlyLab handoff (v0.6.1, 2026-09-24) Read this first if you are Claude, GPT, or another agent continuing the repo. Owner: https://github.com/pinkysworld/FlyLab -Package version: **0.6.0** in `CITATION.cff` and in the paper. **`flylab/__init__.py` and `pyproject.toml` both still say `0.5.0`** — see Issues; those files are not mine to edit. +Package version: **0.6.0** in `CITATION.cff`, `flylab/__init__.py`, `pyproject.toml` and the paper. ## What the project is @@ -64,7 +64,7 @@ flylab serve # http://127.0.0.1:8765 Read these before writing any new claim into the paper, the README or a PR description. -1. **The dependence verdict is a joint property of the prediction and the extract, and recurrence predicts it. That is the headline.** The same landscape, same instrument, no parameter changed: on the 1-hop `named` cut (mean degree 1.208, 85% of edges onto four seed cells, 184 of 1126 nodes with any input) 11 of 84 cells are topology-dependent, 42 composition-dominated and 31 no-effect at n = 1000 with Benjamini-Hochberg and a 1 % relative effect floor. On `taste_motor` (mean degree 10.4) the same cells give 54 / 0 / 30. The **scaling study** (`papers/scale_study.json`, T27, `flylab.analysis.scale`) then settles what the inversion tracks: over seven extracts the composition-dominated verdict survives on **exactly one**, the in-star, and `scale_1k` — *fewer* nodes than `named` (1000 vs 1126) but mean degree 22.9 — is already topology-dependent. **Recurrence, not size.** At 5k and 10k imidacloprid is distinguishable from *every* null; the top rungs run at n = 20 (resolution 0.048) and so confirm rather than establish. "Most predictions do not need the connectome" and "the topology-dependent set is exactly the chloride-channel blockers" are **withdrawn**. Never state a dependence result without the extract, its structure, the compound, the concentration, the readout, the *p* and the correction. +1. **The dependence verdict is a joint property of the prediction and the extract.** The saved comparisons are consistent with recurrence tracking the verdict, but they do not isolate a causal structural feature. On the 1-hop `named` cut (mean degree 1.208, 85% of edges onto four seed cells, 184 of 1126 nodes with any input) 11 of 84 cells are topology-dependent, 42 composition-dominated and 31 no-effect at n = 1000 with Benjamini-Hochberg and a 1 % relative effect floor. On `taste_motor` (mean degree 10.4) the same cells give 54 / 0 / 30. The **saved scaling study** (`papers/scale_study.json`, T27, `flylab.analysis.scale`) shows the composition-dominated verdict on **exactly one** of seven extracts, the in-star; `scale_1k` has *fewer* nodes than `named` (1000 vs 1126), mean degree 22.9, and is already topology-dependent. At 5k and 10k imidacloprid is distinguishable from all three ranked edge/wiring nulls; at 10k the separate plain transmitter-label permutation is indeterminate. At 25k and 50k the three ranked wiring nulls remain distinguishable, while the transmitter-label checks are not both distinguishable. Those rungs run at n = 20 (resolution 0.048) and so confirm rather than establish. The larger cut files and source Feathers are absent from the versioned checkout. A separate 2026-09-24 exploratory follow-up rebuilt the cuts from public v1.0 inputs and reran both confirmatory compounds at n = 100 (5k) and n = 50 (10k–50k); its eight profiles all classify topology-dependent, but n < 1000 and no across-profile correction was applied. See `data/derived/scale_study_high_power_2026-09-24.json`. The committed 1k node/edge lists match exactly and the larger cuts match T27's counts, but historical byte identity is unverified because those hashes were not recorded. "Most predictions do not need the connectome" and "the topology-dependent set is exactly the chloride-channel blockers" are **withdrawn**. Never state a dependence result without the extract, its structure, the compound, the concentration, the readout, the *p* and the correction. 1a. **Non-rejection is not equivalence.** Each mode carries a three-way verdict against a prespecified margin (0.332 Hz, 5% of the vehicle readout on `named`). For imidacloprid at n = 1000: equivalent within tolerance for rewire_degree_preserving, sign_permute_weight_matched, weight_permute, indeterminate for sign_permute. Never write "reproduces". 1b. **The transmitter null was biased.** Plain label permutation moves the cholinergic share of out-weight from 0.619 to 0.544 ± 0.027, putting the real graph at the 100th percentile of its own null. It is now reported as a joint target-set-and-sign null; the ladder's rank-3 rung is the weight-matched shuffle. 1c. **The instrument is validated.** Planted ground truth on a synthetic cut: the ladder recovered 2 of 3 planted topology-dependent effects at n = 200 and returned no false positive on the unplanted control. False-positive rate 0 of 18; detection by permutation count n=50: 0.61, n=200: 0.61, n=1000: 0.67, so effect size dominates. @@ -83,8 +83,8 @@ Read these before writing any new claim into the paper, the README or a PR descr ## Issues - #1 Gate 0 occupancy CLI — **closed**. #2 Gate 1 map-extracted taste GRNs — **closed in v0.5**. #3 Gate 2 five-drug panel — **closed**. -- **Open: version strings disagree.** `flylab/__init__.py` and `pyproject.toml` both report `0.5.0`; `CITATION.cff` and the manuscript say `0.6.0`. Pick one and bump the two package files. Not fixed here because those files belong to other owners. -- **Open: expression coverage is 0.205** and adult motor-neuron receptor expression is a confirmed literature gap — MN9, the main readout, carries the `unknown` default weight. +- **Version strings are aligned at `0.6.0`.** `flylab/__init__.py`, `pyproject.toml`, `CITATION.cff` and the manuscript currently agree. +- **Open: receptor annotation and MN9 expression.** The 0.205 value is the share of graph-node × receptor-key pairs with a coarse class-level annotation across the two committed cuts, not a share of cells measured to express receptors. GEO GSE141807 is now reprocessed into VNC-wide gene-detection fractions after published QC, but no cell-class labels are joined; see `data/literature/vnc_receptor_detection_GSE141807.yaml`. A two-cell discrepancy in one replicate versus the paper is recorded and unresolved. The dataset does not identify MN9, whose proboscis motor-neuron arbors are in the brain subesophageal zone; MN9 still carries the `unknown` default weight. - **Open (new): `flylab/analysis/nullmodels.py` still exposes `connectome_information_score`**, a v0.5 summary superseded by the dependence class and the necessary information level. The paper no longer quotes it. Either delete it or document it as deprecated. - **Open (new): docstrings in `flylab/` still say "occupancy" in places where the value is a functional engagement** (e.g. `flylab/pharm/occupancy.py` module name and several assay docstrings, `hill_occupancy` kept as a deprecated alias). The behaviour is correct and typed; the naming lags the paper. - **Open (new): `flylab/analysis/predictions.py` still labels H1–H7 "pre-registered"** in its docstring and output fields. The paper now says *prospective*. Rename when convenient. @@ -158,10 +158,11 @@ python scripts/reproduce_paper.py --list ## Next work, in order -1. **v0.7 gate 7 — expression coverage.** Either source adult motor-neuron receptor expression or declare the 0.205 coverage a permanent bound. A data problem, not a modelling one. +0. **Review the scale follow-up.** `data/derived/scale_study_high_power_2026-09-24.json` has an exploratory higher-power rerun, but it is not a replacement for T27 or confirmatory evidence. The scheduled ladder step is now conditioned to run on schedule as well as manual dispatch; after the workflow change reaches the remote branch, verify its uploaded artifact and preserve input/cut hashes. Any future higher n should be budgeted from measured runtimes, with the same simple degree-preserving graph and seeded permutation procedure. +1. **v0.7 gate 7 — receptor-expression evidence.** Determine whether adult VNC motor-neuron labels can be joined defensibly to the reprocessed GEO GSE141807 barcodes; if so, report class-specific detection with annotations, sample/sex composition, detection rule and dropout treatment. Resolve MN9 separately with a class-matched brain/SEZ atlas or targeted assay. Keep gains uniform unless this produces defensible cell-class evidence. 2. **The one allowed fit.** The uncertainty budget names it: the engagement→gain transformation. Calibrate it against evoked postsynaptic responses at a known synapse; do not fit a potency value. 3. **The free win.** Re-run the dependence and selectivity analyses against synapse-confidence strata: `weight_threshold` carries S1 0.294 and costs no bench time. 4. **One live assay.** Climbing first: a full published protocol and control statistics exist. Import by hand into `live_lab`; nothing else may write it. -5. **Housekeeping:** reconcile the version strings, mint the Zenodo DOI, and only then consider the predictions pre-registered. +5. **Housekeeping:** mint the Zenodo DOI; predictions remain prospective until a tagged release is externally archived. **Do not** start a 166k-cell LIF, add a second connectome, invent live PER/climbing numbers, claim that no tool has combined connectomes and pharmacology (FlyBrainLab and the receptor-map whole-brain models exist and are cited), or accept a PR that only adds a prettier viewer. diff --git a/README.md b/README.md index fd6c286..ef0c9a1 100644 --- a/README.md +++ b/README.md @@ -47,6 +47,8 @@ dependence | ablation | robustness | uncertainty The result is not just a simulated firing rate. FlyLab records what evidence entered the model, which assumptions transformed it, which graph was used, and how strongly the prediction depends on those choices. +The browser bench saves your primary run and experiment-design settings in the current browser. Use **Export setup** and **Import setup** to share those settings as a versioned JSON preset; the preset contains inputs, not calculated results. On narrow screens, the two tab rows collapse into one grouped panel picker. + ## Why it is different ### 1. Evidence is typed, not flattened into one fake "EC50" field @@ -72,7 +74,7 @@ This makes it possible to separate effects that are sensitive to detailed wiring The ladder itself is validated rather than assumed to work. A recurrent loop of known strength is planted in a synthetic graph of the same size, density and transmitter composition as a real cut, and the analysis is asked to find it; the unplanted control gives an empirical false-positive rate, and a power surface maps detection against effect size and permutation count. -**The verdict this analysis returns depends on the extract it is measured on, and recurrence — not size — predicts which way.** The repository's two committed cuts disagree: the same landscape classifies most cells composition-dominated on the sparse 1-hop `named` cut and none of them on the denser `taste_motor` cut. Run across a ladder of nested cuts of 1k to 50k cells (`flylab.analysis.scale`), the composition-dominated verdict survives on exactly one extract — the in-star — while a cut with *fewer* nodes but twenty times the mean degree is already topology-dependent. A dependence result is therefore only quotable together with the structure of the graph it was measured on. +**The verdict this analysis returns depends on the extract it is measured on.** The saved scale study associates the change more with recurrence than node count, but it does not isolate a causal structural feature. The repository's two committed cuts disagree: the same landscape classifies most cells composition-dominated on the sparse 1-hop `named` cut and none of them on the denser `taste_motor` cut. Across the saved 1k–50k ladder, only the `named` in-star is composition-dominated; `scale_1k` has fewer nodes but twenty times its mean degree and is already topology-dependent. At 5k and 10k the imidacloprid effect is distinguishable from all three ranked edge/wiring nulls; transmitter-label shuffles are separate and indeterminate at some rungs. The larger scale cut files are absent from this checkout, so those results remain imported observations. Quote any dependence result with the structure of the graph it was measured on. ### 3. Model conclusions are stress-tested @@ -86,7 +88,7 @@ FlyLab includes: | **Global uncertainty** | Which assumptions dominate variance in a model output? | | **Value of information** | Which measurement or re-analysis would reduce the most model uncertainty? | | **Claim provenance** | Which links in a result are observations, literature-derived inputs, modelling assumptions, or computations? | -| **Scale ladder** | Does the dependence verdict settle as the extract grows, and what structural property predicts it? | +| **Scale ladder** | How do dependence verdicts vary across saved extracts of different sizes and structure? | Negative results are first-class outputs. If a conclusion depends on an asserted gain rule, or a connectome-specific interpretation is not supported by the null analysis, FlyLab is designed to show that rather than hide it. diff --git a/ROADMAP.md b/ROADMAP.md index 3d66f14..36a969d 100644 --- a/ROADMAP.md +++ b/ROADMAP.md @@ -30,19 +30,20 @@ Full research plan: `docs/RESEARCH_MAP.md`. Numbers: `papers/results.json` (rege | **6f Subunit-resolved nicotinic keys** | α6 (spinosad) and β1 (imidacloprid) split out; the rest stay on a documented aggregate | **done in v0.6** | | 6g IJRC research article | Intro / Methods / Results / **Discussion** / Conclusion, inside the word target | **done in v0.6** — drafted, not submitted | | **6h Second-round referee revision** | weight-matched transmitter null, three-way equivalence verdicts, FDR with confirmatory/exploratory labelling, a relative effect floor, a direction-aware ablation baseline with a matched reference distribution, the engine's row normalisation documented and swept, the transmitter sign table published, ground-truth validation of the dependence ladder, and the landscape repeated on the denser cut | **done in v0.6.1** — and it **reversed** the central RQ2 claim | -| 6i Scale ladder | the same profile across nested extracts of 1k-50k cells, with the permutation budget scaled to the edge count | **done in v0.6.1** — the verdict settles as topology-dependent; `scale_study.json`, T27 | -| 7 Expression coverage | adult motor-neuron receptor expression, or a declared permanent bound | **open** — coverage 0.205, MN9 unresolved | +| 6i Scale ladder | the same profile across nested extracts of 1k-50k cells, with the permutation budget scaled to edge count | **saved result in v0.6.1** — T27; exact 5k-50k cuts are not present in this checkout, so those rungs cannot currently be rebuilt | +| 7 Expression evidence | map adult VNC motor-neuron labels to reprocessed counts and resolve MN9 separately | **open** — GSE141807 now has VNC-wide gene-detection fractions; cell classes and MN9 remain unresolved | +| 7b Scale power | rerun exact 5k-50k cuts at higher permutation resolution | **gated** — cuts missing here; profile/optimize exact edge-swap cost before raising n | | 8 Live fly | one PER or climbing table in `live_lab` | **blocked — protocol only** | | 9 Archival | Zenodo deposit; only then may "prospective" become "pre-registered" | **open** | ## What v0.6.1 changed about the plan -The dependence verdict turned out to be **substrate-dependent**, and the scaling study then said what it tracks. Across seven extracts the composition-dominated verdict survives on exactly one — the in-star `named` cut — while `scale_1k`, which has *fewer* nodes (1000 vs 1126) and twenty times the mean degree, is already topology-dependent. **Recurrence, not size, predicts the inversion.** The generalisation "most predictions do not need the connectome" is withdrawn. Consequences for the roadmap: (a) no dependence result may be quoted without the structural statistics of the extract it was measured on (`flylab.analysis.dependence.cut_census`, T19, T27), (b) the permutation budget must fall with scale, so the top rungs (n = 20, resolution 0.048) confirm rather than establish a verdict, and (c) the paper is now framed as a computing contribution with the fly pharmacology as its demonstration domain. +The dependence verdict turned out to be **substrate-dependent**. In the saved scale study the composition-dominated verdict occurs on exactly one of seven extracts — the in-star `named` cut — while `scale_1k`, with *fewer* nodes (1000 vs 1126) and twenty times the mean degree, is already topology-dependent. This is consistent with recurrence tracking the change, but the study does not isolate a causal feature. The 5k–50k source cuts are absent from this checkout, so T27 cannot be independently rebuilt here. The generalisation "most predictions do not need the connectome" is withdrawn. Consequences for the roadmap: (a) no dependence result may be quoted without the structural statistics of the extract it was measured on (`flylab.analysis.dependence.cut_census`, T19, T27), (b) the top rungs run at n = 20 (resolution 0.048) and confirm rather than establish a verdict, and (c) the paper is framed as a computing contribution with the fly pharmacology as its demonstration domain. ## v0.7, in order -0. **A better-powered top of the ladder.** The scale ladder settles the verdict, but its top rungs run at n = 20 (resolution 0.048), where a rejection is the smallest the test can express. More permutations at 25k-50k, or a cheaper null, would turn confirmation into evidence. -1. **Expression coverage (gate 7).** Either source per-cell-type receptor expression for adult leg/labellar motor neurons or declare the 0.205 coverage a permanent bound. A data problem, not a modelling one. +0. **A better-powered top of the ladder.** T27 is an imported saved result; the exact 5k-50k cuts are absent from this checkout. Restore those exact cuts, profile the degree-preserving edge swaps, then increase n only after the optimized sampler preserves degree, simplicity, seeded reproducibility and the same permutation test. The existing 25k-50k results use n = 20 (resolution 0.048), where a rejection is the smallest the test can express. +1. **Expression evidence (gate 7).** Determine whether adult VNC motor-neuron labels can be joined to the reprocessed GEO GSE141807 barcodes and, if so, report class-specific detection with pinned annotations, sample/sex composition, detection rule and dropout treatment. The existing counts are VNC-wide and do not identify MN9: it is a proboscis motor neuron in the brain subesophageal zone and needs a separate class-matched atlas or targeted assay. Keep uniform gains until evidence supports a change. 2. **The one allowed fit.** The uncertainty budget now names it without ambiguity: the engagement→gain transformation (S1 0.410, VOI 1.29 of 3.149 Hz²), not a potency value, which the model cannot see at a saturating dose. 3. **The free win.** `weight_threshold` carries S1 0.294 and needs **no experiment**: re-run the dependence and selectivity analyses against synapse-confidence strata of data already held. 4. **One live assay (gate 8).** Climbing first — a full published protocol and control statistics exist; the PER baseline could not be sourced and is recorded as `null`. diff --git a/artifact-manifest.json b/artifact-manifest.json index f58d6f4..30fe355 100644 --- a/artifact-manifest.json +++ b/artifact-manifest.json @@ -7,9 +7,9 @@ "paper" ], "flylab_manifest_version": "1.0", - "flylab_version": "0.6.0", - "generated_utc": "2026-09-20T16:42:40+00:00", - "git_sha": "b82b8edaaabbab7be1f2b6aec639dd8a076feb27", + "flylab_version": null, + "generated_utc": "2026-09-24T17:56:28+00:00", + "git_sha": "50f5465b2df028a6cb2b93926cfbd14d7c746c9b", "git_tracked": true, "kind": "flylab-artifact-manifest", "note": "Hashes of the inputs the committed results depend on. Regenerate with `flylab manifest --write` whenever code or data changes, and before tagging a release: `flylab manifest --check` is what the release workflow gates on.", @@ -19,19 +19,19 @@ "entries": { "flylab/__init__.py": "7080f58cb9c2fef8f8f0680c37a61bd4503726c43de73366087991604f131616", "flylab/analysis/__init__.py": "a6029eeec0258ce9799bde5f62b00fa03452c886abdb9881d38505f3f576ab1b", - "flylab/analysis/baselines.py": "600093fb62a79d0dfe21077649cbda8016e4e9d30139638582936cf72c127939", - "flylab/analysis/claims.py": "d657444619f3c4218b8224ad09df140494a504b50b25fabc8aa118d3a3e6e391", - "flylab/analysis/dependence.py": "6b6a98ea8e008318509e68e6a513d3bd162d7e7f173d13bcd41b8968043ce9d7", + "flylab/analysis/baselines.py": "65a674328fe92c2ad00fd7f44019ee513130cca2c4430da02ee8d8fde580b939", + "flylab/analysis/claims.py": "dbe948f27a7827bb75a9f116b54e551ccfc5706e450db3b63fd8118ea9d543ed", + "flylab/analysis/dependence.py": "b20afd704c5a94c3500965e34f21dc33f0c72e833e24553566190273cd029047", "flylab/analysis/fit.py": "b6946fffa209c7444a054198e462a1f43cdf0dd0d12c084803579e2ef48f0ddf", "flylab/analysis/impact.py": "72f68e33369c92f3e47e957e7d50294941ba931441e56d1486f59a80f167dcd6", "flylab/analysis/layout.py": "771f168fac412e6747e98ae0f8bf0a39221a07694a41fa641e4b84a226f9a773", - "flylab/analysis/nullmodels.py": "53f81c146483d26d5ec8649d43f3e483560d8f856d1bb30a6ec13281da9ddb9e", + "flylab/analysis/nullmodels.py": "5f93a5a47945ebe412077d006decc8cc538ca3c2bccb55e48609377a08be48e0", "flylab/analysis/predictions.py": "3382c24628274db35d60199a0be428b89a890efce250e944e2651245c1105e83", "flylab/analysis/robustness.py": "31978a1371fb0863793bd0c4e6191ff201b1582b901de6c334f838b30857ec8e", "flylab/analysis/scale.py": "6264d2ce3497b1cbc075617b3d2ec4f532e092e5b8646aa665cdd76d6615f581", "flylab/analysis/selectivity.py": "622c881f58d50806a2f344a1c721a2a07dacc896e6718210fedbfc01250f2ef6", "flylab/analysis/uncertainty_global.py": "6adab31073fa0b7f933c2054af9e70f070abc0a69e641626f48421c6e0ae2d02", - "flylab/analysis/voi.py": "866fdf95787d61fd96abda825418a1b78633894bb506a0c90f691653c04cbab8", + "flylab/analysis/voi.py": "2445b6944e3190fb264d09fae0f61492b0aafb171a192fdffdae2a6bceb22f27", "flylab/assays/__init__.py": "6f69d06efa5d2db4f6cbcb83c2a8124184bb1cf299c6dd4509d8427aa876fad4", "flylab/assays/ensemble.py": "664b12ff7e0f7864e90d167b899573727d5aec9c729bb70f49cbef795cbb725e", "flylab/assays/experiment.py": "89356ddc84c94d785604e735d4e3873deeb2601c6b18346949eb75d342999277", @@ -42,12 +42,12 @@ "flylab/assays/taste_map.py": "189bb330536a4a706a4fec665c01b9c569b42f22cf1391dff825eb14095d34be", "flylab/assays/wholens.py": "9d771857e6780f2b6f5d17de8cff3111b39e4106bcbce0decbb46db289be4657", "flylab/browser/__init__.py": "8a2e74bb1b286e5277745ce1bff2f82b40c673f20025a521d8207c3d3025fa26", - "flylab/browser/bridge.py": "fa8dc01c2b97282fe2c3f9c28d7cfb63751d5d77e19c428fa19640cc24bb3f0d", + "flylab/browser/bridge.py": "f116c2c9fed2e95c4a8f67580017023fa12b24f70de50f6d66da45b830d7b2ac", "flylab/circuit/__init__.py": "0b5f006f027f8caa671f34334a898d9ff640e8607f9814a606b735cfa423028d", "flylab/circuit/lif.py": "13bab4a5156d79537147acea5d893d069faf802a6526d5f95ff3d1fd400db9bd", "flylab/circuit/rate.py": "3f47664856d6fa56146dfab6c900082ea847ce81cad8d88f5317c47f2e2dd3f3", "flylab/circuit/reduced_taste.py": "2db7d2fa0b9d2413ca6593fcbd3d6e837a7ba33b84ca593bb98da2e39d277ca8", - "flylab/cli.py": "4d75d90546e8f7a1e780496f7ff872b39bf9a150e8894dc1cc8c0882b988b81a", + "flylab/cli.py": "555d203c15d64137ed3e6bcd0f7043c78991701ff9746c4783fb8c1a04217dbb", "flylab/maps/__init__.py": "01b46035d1cf4ca1e60115200fe04e4dc44d4f5a86c67a5d80c85ebd1b92d37c", "flylab/maps/extract.py": "c731fa9329c32fc380c641c0d0b0f376368d1d8eb554597f9f2e97f74e307e78", "flylab/maps/ladder.py": "29ace69ee0b0e58c4b8b5be49b53402e5ef17e23e22f97f169b96cc726ff6cc7", @@ -58,7 +58,7 @@ "flylab/pharm/binding.py": "61a05b32a4939ff747794d4e68083e5192cf49aff9f6784380439efc0253cd82", "flylab/pharm/evidence.py": "d566d0cc95c5fb4a522bf1c9c16e70f998d4be44f6d15abc7602d355dd797a50", "flylab/pharm/exposure.py": "03b8b217375d918099318d934f2e356440e4bf39d26938571829b760d811f156", - "flylab/pharm/expression.py": "b855ad8277af09d75a35f13f6d418d5d4c2d0f71dbdea0f683f8f87abcef692c", + "flylab/pharm/expression.py": "3242b77be02ddb72b1189a951e9bb53417d544121b2ee1d177b1bb77c647b6bc", "flylab/pharm/genotype.py": "69d93db35db9c8f352e538d8f8b365de4712b36dc96a3cd5444760fa4ddb46ae", "flylab/pharm/mechanisms.py": "2c1377cbcff26ace97cbab917721996fc0416c939b27d9772bf4a69f3f4fd3ff", "flylab/pharm/mixtures.py": "3cc56f8e39adfdb41eeb8300c88b87e370320597ab47db3dc89d31c071cf2535", @@ -72,60 +72,61 @@ "flylab/validation/rank.py": "d82c9590d10943de082d12dab91ccbea5cad8742c9a7fb5376ebf67bb24d5292", "flylab/viz/__init__.py": "eabed1cfef125f0af50befc1e046e1270ad3339cb3aa0e1936aa3fd31c137a6d", "flylab/viz/figures.py": "d3c4370400f4eb2b38bf6dfd0459529637e2763dfe96a9b17799f54397e03b07", - "pyproject.toml": "eab14bbee64da107f20259927bd9325c29e376286fef569423c50cc02b127140", + "pyproject.toml": "9209766857017288802b647982d968c470633f141a3fdf8f1019af13d66fd04b", "scripts/__init__.py": "8e742cba903af8699473573f6c8253b581f30fe9d14e6162e7f74c5ac1b6c5fa", "scripts/build_pages.py": "56185a50fd8d69736d6433f0eb4b0d26b032e0da0b88b3ceee3a5240c7311e34", + "scripts/check_commit_attribution.py": "63f90c7d3bd4ef5ee9dbd5bfc4197fc51500fd92d1e5e34f0498e7f4a66316bc", "scripts/manifest.py": "b73a4954d86d3b49c4cc2033bf2c9d0dcba31420a2a36fe77b8743a7f6927fc6", - "scripts/reproduce_paper.py": "45d0cd592a3e24971593c2587601aa4afd6e5e48c0ce8f639ba5ed38f5007f42" + "scripts/reanalyze_vnc_expression.py": "b9e3003fdf5bd245d544bc46dcbc76dc425cae335468e2c2ec79afa5d426da2a", + "scripts/reproduce_paper.py": "b49fa50da9764ad51429b82a5994af1723cbb91f20ee9d81bf56987cfab31f6d" }, "fatal": true, "kind": "files", - "n_entries": 60, + "n_entries": 62, "patterns": [ "flylab/**/*.py", "scripts/*.py", "pyproject.toml" ], - "sha256": "4a40094ca024eede4b8d5cdf540459c7ec1f9381218411ec9f82d44488c9535f" + "sha256": "70a7d0d3e47fc27622452576b4a34d2eb8427fb1df6e7235bec1da2591c77189" }, "dependencies": { "description": "resolved versions of the project's dependency closure", "entries": { - "PyYAML": "6.0.1", + "PyYAML": "6.0.3", "Pygments": "2.21.0", "annotated-doc": "0.0.5", "annotated-types": "0.8.0", "anyio": "4.15.1", - "certifi": "2026.2.25", + "certifi": "2026.7.22", "click": "8.5.0", - "colorama": "0.4.6", "contourpy": "1.3.3", "cycler": "0.12.1", "fastapi": "0.141.1", - "fonttools": "4.65.0", + "fonttools": "4.61.1", "h11": "0.16.0", "httpcore": "1.0.9", "httpx": "0.28.1", - "idna": "3.11", + "idna": "3.20", "iniconfig": "2.3.0", - "kiwisolver": "1.5.1", + "kiwisolver": "1.4.9", "markdown-it-py": "4.2.0", - "matplotlib": "3.11.2", + "matplotlib": "3.10.8", "mdurl": "0.1.2", - "numpy": "2.4.6", - "packaging": "24.0", + "numpy": "2.4.2", + "packaging": "26.0", "pandas": "3.0.6", - "pillow": "12.3.0", + "pillow": "12.1.1", "pluggy": "1.6.0", "pyarrow": "25.0.1", "pydantic": "2.13.5", "pydantic_core": "2.46.5", - "pyparsing": "3.1.1", + "pyparsing": "3.3.2", "pytest": "9.1.1", "python-dateutil": "2.9.0.post0", "rich": "15.0.0", "shellingham": "1.5.4", - "six": "1.16.0", + "six": "1.17.0", "starlette": "1.6.0", "typer": "0.27.2", "typing-inspection": "0.4.4", @@ -135,11 +136,13 @@ "fatal": false, "kind": "packages", "missing": [ + "colorama", "exceptiongroup", + "playwright", "tomli", "tzdata" ], - "n_entries": 40, + "n_entries": 39, "note": "Informational: a manifest is verified on machines other than the one that wrote it, so a version difference is reported and never fails a check.", "patterns": [], "python": "3.11.15", @@ -149,6 +152,7 @@ "matplotlib", "numpy", "pandas", + "playwright", "pyarrow", "pydantic", "pytest", @@ -156,29 +160,30 @@ "typer", "uvicorn" ], - "sha256": "60a916447ed6449fa0c371058ca74bd68ae97df26bdf3a1517eaf8ded595ddf5" + "sha256": "c39b913e4c05eaa8d6e4d56bbd008f6554d6555330fad94a993f8f04a9a473b7" }, "graphs": { "description": "the derived MaleCNS cuts; CI products, never hand-edited", "entries": { "data/derived/malecns_census_v1.json": "7aebb432cea83eebce8f4f6a2d0967a79f782bad6bee20c05898cfc1526479f9", "data/derived/malecns_gustatory_seeds.json": "616adbe67df4034b4f1e0b1a3d363fd46c5f5a1b9e402bb44699cec6ae783b03", - "data/derived/malecns_named_neighborhood.json": "7f296ec5ec19490c46560732be83119a872668a87b70932e07aecf64443f49ae", + "data/derived/malecns_named_neighborhood.json": "8ae12b354372d1a4fcc428cca3eab88ea72b7027e8108fcf7abbaa966c0c63d1", "data/derived/malecns_scale_1k.json": "3ff5224ec586a9da2a9ea9ab007736c357ca76324f52d8f953e4085cb21437a9", - "data/derived/malecns_taste_motor_neighborhood.json": "e4a28229b2c469a39a962aac15a01693effd9dd18d65694968be39c216cc0473" + "data/derived/malecns_taste_motor_neighborhood.json": "e4a28229b2c469a39a962aac15a01693effd9dd18d65694968be39c216cc0473", + "data/derived/scale_study_high_power_2026-09-24.json": "69a36e3162f5d510dbfcfe382c851bf537ee8327db1af213776dcfbc832c1365" }, "fatal": true, "kind": "files", - "n_entries": 5, + "n_entries": 6, "patterns": [ "data/derived/*.json" ], - "sha256": "f7bc155e879a17c6a15349c746f5f4439ee357d9d35b0dfb5bacb2deeffb9f2d" + "sha256": "1569948f8c5785f343840018544fc8e8cec22cef24b59397c800e37f728a3a80" }, "library": { "description": "the typed compound library (schema v3)", "entries": { - "flylab/pharm/library.yaml": "594b8d32c5e61ebd4de31cc77441aeb82503a0275a7a6ef951ec7a86a48732a8" + "flylab/pharm/library.yaml": "359096467b33242174dcc5d787a6b7fc6d37573f8daf4771efd159b64ca95e05" }, "fatal": true, "kind": "files", @@ -186,113 +191,114 @@ "patterns": [ "flylab/pharm/library.yaml" ], - "sha256": "7cc562e777234a683cb77c203dc13b4e58240cbdb52eaeb98f06a9c4a6c2e4e7" + "sha256": "845e3a529220ebf83b537930d0a9bb44b05329065239dcbe9a19743fc53dd753" }, "literature": { "description": "the sourced literature datasets", "entries": { "data/literature/behavioral_assays.yaml": "c8e8043136456d817cd9985f0f12a293fbb616a9620dc5172066957238ae952d", - "data/literature/fly_pharmacokinetics.yaml": "a44b6f86a79e9163b43858822a7a3888e9eb19b07825cdf08652c584430bf0c9", + "data/literature/fly_pharmacokinetics.yaml": "a4099490f9675373da9629448cd1fc2fdbf71c316f95ab41e0d1765f457ba024", "data/literature/mixtures.yaml": "ad569d14b0904a344bbd22cfb44f8886f9c5f9cc7de38c70312e8a308386a0fb", "data/literature/published_rank_orders.yaml": "192c884350ec86ff94d235082e0b8e2d728ac06dc751ddab601bc63890023123", - "data/literature/receptor_expression_by_class.yaml": "dac670f9a587f7797eec8ced8fdb3a03dc11d1a57432a4ffd68a9e3388b42727", - "data/literature/resistance_alleles.yaml": "8d991ceb691fd3b449d6ddfb06955c1c7f102e599b551ff99e9525f1801db973" + "data/literature/receptor_expression_by_class.yaml": "3d8c104cbc27445c4e32c2ec606d34e944dad49103ecae1673fe33ff17fc98a0", + "data/literature/resistance_alleles.yaml": "8d991ceb691fd3b449d6ddfb06955c1c7f102e599b551ff99e9525f1801db973", + "data/literature/vnc_receptor_detection_GSE141807.yaml": "8f494d433cbdd6a6a06e3023af2c4fbc109a85285098c9242893c2b9bac43628" }, "fatal": true, "kind": "files", - "n_entries": 6, + "n_entries": 7, "patterns": [ "data/literature/*.yaml" ], - "sha256": "8a5001659a335ce180e1a22313c955faf036947516b56c7f779194c2262c4a1f" + "sha256": "c2efaf78ab8acf97e96833115d043320a9d5f260b030d05fb7749e4460767116" }, "paper": { "description": "the committed paper artifacts regenerated by scripts/reproduce_paper.py", "entries": { - "papers/IJRC_FlyLab_draft.md": "640671d5c83bdb58fbd4782a81fcf230bc6b96a42f62716fdd9bf6feba26c606", - "papers/STATUS.md": "b9024f67ddc100a5e6683510b4caa9443b17f2666f780e1975ebd79760999c67", - "papers/SUBMISSION_CHECKLIST.md": "b54b5a6e38ceb3e7a4b810544d8606fe6e52fce7a358fabf08a1e18f767008f6", - "papers/SUPPLEMENT.md": "ff391afbe3a3651cb510ed82060f63264611c426cb5c8f3a045b0eda9f92ce4d", - "papers/figures/F10_rank_validation.png": "4943edf7eedeccebd5d8c642a3629add43997ece4ad83b19a634f11c7abb1222", - "papers/figures/F11_evidence_types.png": "71c27c1084780bd0d502d9584386ba006aaab28a74bfdbfc690d4dbb8b2b7287", - "papers/figures/F12_dependence_landscape.png": "2dc05bbb7a2fc49b61a4481ed5c9984927583d747caaa43c3403791977438c89", - "papers/figures/F13_ablation.png": "f526c9d315fc0be13ef8216ce5ab824c8001ad3486fef8a6454d302cf4ed5560", - "papers/figures/F14_conclusion_stability.png": "f4a8474416dc694c703696343717436ae3b2ad39abcdb23142fbb39b0510c2f2", - "papers/figures/F15_uncertainty_voi.png": "17855019835772b11757de863b8bae71d200524d4ec39f3508278f1029f2bc4b", - "papers/figures/F16_instrument_validation.png": "7cc64976c52d975dfbd3c25d0f9279cdd6b3ce744709cbd8cc819253bf3657fc", - "papers/figures/F1_architecture.png": "649251040d5e94e209901c6da13eb96715496b458a72d2bcf7858f30006b324b", - "papers/figures/F1_architecture.svg": "f499b2a41f2016a74b4424b37e75e587fdf80bdfa283ee9a158ff0580850da87", - "papers/figures/F2_dual_scorecard.png": "22e67e15642afc6ef30a31d07c28eb663d0b9b2b013567cf741f888940e1b3df", - "papers/figures/F3_occupancy_curves.png": "45ed0c6f4f3107ff350ddeb4808947c145556a6ba74bf8f1f970094d7b5e5b0b", - "papers/figures/F4_taste_motor_graph.png": "7474e1a85dcadcc4296b77912f517b5635fc034ea58d63511a97c8c072d1c319", - "papers/figures/F5_bitter_veto.png": "879920de976d9ac6d0ebe402c20cf0fedc2dea9a69b1fe7d339c751f665f2dc2", - "papers/figures/F6_dependence_profile.png": "838219d044a94200adbf6b04270c510a4ec3b3e625bb23b262c2fd4e8b942918", - "papers/figures/F7_selectivity_landscape.png": "57d091db0d150e8e11d472b8ac5f5d283d0efe6706f742e330c3c1ea2436424d", - "papers/figures/F8_dose_response_sensitivity.png": "f014b43b16ab8d1324604d0a1e8db1581f39c8dd9b102e7428be237e1ba190e4", - "papers/figures/F9_genotype_panel.png": "f51b3f7a33c7b949b98c35cd2d40d4dc4606aca5a4dd7262b59da0d27f56f3fd", - "papers/figures/captions.md": "697969d5df58bf2bce1eed8f33a137b9ca5d90013d51aab0858da53d7dede7bb", - "papers/results.json": "fab4a68fce4a1e3fd2aa472485191c492511e4a8c149660425efdd80ddb86513", + "papers/IJRC_FlyLab_draft.md": "2f5e69fd6715b2d44f17516dbe6f771197a5c71b1fba6a8240ef33678a66c5e7", + "papers/STATUS.md": "20561f3482c603f4bd4fe0addb9979493d57a5de72e5be6c781dd946f27499c3", + "papers/SUBMISSION_CHECKLIST.md": "3d487ea123372ace6a8972a8355a3d2785d394584f5dad07edcb31597959b181", + "papers/SUPPLEMENT.md": "adf7b60449cee9ac467fb63c3da37b61b0fa9a26f5aaf031acf623a9a94c03a2", + "papers/figures/F10_rank_validation.png": "196cf70a0c73e33e4197bef66f6ffd8e2b8528e0bf216811dfd1b1e0807acda4", + "papers/figures/F11_evidence_types.png": "9fb72e6402152cc0a8e0a16b491dd6b3459cd60af2a18a6d2fcb4dab87adc281", + "papers/figures/F12_dependence_landscape.png": "538a2e99cdd2e2185098ce82807698d7d9745631a4e85b54b4d1e050c67057e2", + "papers/figures/F13_ablation.png": "75921fb8d60c32c9c9039d55578503e3a675fd18ae77a95978ba93709651a886", + "papers/figures/F14_conclusion_stability.png": "60833864c4cbf63ed9eeb0c1dd29d11438809312abb6486c6de095aa6ec71c27", + "papers/figures/F15_uncertainty_voi.png": "e8fb5749c5337210ac6609140a92df9ce1ef9b6b18303f4599cd1834990cb575", + "papers/figures/F16_instrument_validation.png": "070f9f745f51f8be8e96ed4603508f35bf7624a85931042555bd309bb3f10583", + "papers/figures/F1_architecture.png": "3fd437dcbed95933c7ee9479de64610f19add7a362bcc570a1296a3b17ef7c92", + "papers/figures/F1_architecture.svg": "aa8bca97bc88348e60d2be576f9051ed99f5445c89fd47df9697fcc1ca13212f", + "papers/figures/F2_dual_scorecard.png": "60c5421a0828a945b67660251360c0f74097194a9f5f8973e3c26b3045bee6cd", + "papers/figures/F3_occupancy_curves.png": "aa7044de7014697b6b511606ca8cf359c20360627670b30882569a95b8aa4480", + "papers/figures/F4_taste_motor_graph.png": "ca83bf75e41fb919de422d79a865e40f336b4a8fc244abf6b445229fefb6dcc9", + "papers/figures/F5_bitter_veto.png": "b1f6b0a0fb2729301a5c5a3718d865e6f84f2d8ab22f0cb678a88de8e5d4f583", + "papers/figures/F6_dependence_profile.png": "ed3feb5fe15286fa30b234632dde3e63daf1a5ea72c80726332ff2c674bf167b", + "papers/figures/F7_selectivity_landscape.png": "9ddd3bdb7cca7edf4bdabc5dd58bf46fc0238bfba610d1c822981961dc4944a5", + "papers/figures/F8_dose_response_sensitivity.png": "c00bf4ed30cfd4a537addb432326c214a7be6faf6bc0b1b9b9149a34ab781c59", + "papers/figures/F9_genotype_panel.png": "c5aa62e6a22df1c0343e9836470d5ad0fc363bf20c6f85caa4a92e089e498464", + "papers/figures/captions.md": "e27aaa1094ef3d8f4b49c9e50430bf4d9223745c3de02d9ce1b054275df1a74d", + "papers/results.json": "923b2fca85ef3e9bc15945ad4a563f1b4c0739f8656346bb1c60834983796c32", "papers/reviews/round2_referee_report.md": "b58e5c64701ad591959f5b3096a97ab3b1c6f767c3d626c4e0e01b2f3eff44cd", "papers/tables/T0_labellar_types.csv": "d49e71e9dfa9dcc10a38a2d062619d8f834dbec631a2e59d86c73b1f97259298", "papers/tables/T0_labellar_types.md": "3c09c0c2e46e0f54c9542a043604295a208a5705f03b8ee4872b7178aa5ca6e2", - "papers/tables/T10_evidence_types.csv": "d772d1ee5abc3ad54d693ece39b2f1d0df8c1ac9661c6ac350f7d0309c2dc2ad", - "papers/tables/T10_evidence_types.md": "63f7f1570cce9478d8cd5faf3b3aa9e873f64b9985497076de713b6775ea4735", + "papers/tables/T10_evidence_types.csv": "2e26b777ddc0c3e48c169766c947a1fcb4f36e31d08990826b5db4907a71afc1", + "papers/tables/T10_evidence_types.md": "13add02a7331eb7a5a1ee7fc4c91f04d8c7bdc44343cf08cd0151573a9d51f57", "papers/tables/T11_permutation_sweep.csv": "8ff2159495b6e153d502da3af1fe0b4b60a895e6454ed8cf10664d2f0ccf1cd1", "papers/tables/T11_permutation_sweep.md": "d853419ef147ce80ef1081584ce8bb2afcea8af0f0a5ab568b8a326cf6088e51", - "papers/tables/T12_dependence_landscape.csv": "913c269f1011a0c161aa2375b04efa6b7e9f0c0264e9b6c1d33c6f50cc3a60e2", - "papers/tables/T12_dependence_landscape.md": "1c00f1d0d6053d06345f99a5d303620ca8d6a98a32cf898978deb37c0e69f8db", - "papers/tables/T12b_dependence_landscape_taste_motor.csv": "87af2cc0d5da3e390358a8156ce134b3f8f45c1e2e70b24063f662f12a08158b", - "papers/tables/T12b_dependence_landscape_taste_motor.md": "f558db4a5fca7d410fb97637d2e9fc1c9e0ab7445abedf57ad7db27c38f5df17", - "papers/tables/T13_ablation.csv": "68e3fbe9192b84d3398e103f8f9ab37aae56e5538db488dc81e53139df8950c1", - "papers/tables/T13_ablation.md": "826c1aeeca9fde8c3c53884e883fe7f8099a37ff0762e312abbb8bc2dcb1eedb", - "papers/tables/T13b_ablation_information.csv": "5b54afed1d64d30a820d29030864d9bfeb344e3e5f575ee4a243f813642192da", - "papers/tables/T13b_ablation_information.md": "7de9b02ad7f919d8e5c3a1ef3037418cd0138040e3a0402614df1fda8cbaefbd", - "papers/tables/T14_conclusion_stability.csv": "514f1245a9de829445fec6edbf0341fd62e559454541b7bdd35cb7a426bceb7e", - "papers/tables/T14_conclusion_stability.md": "a0bdb6653b01456f079416beda7497e05589091a201c88b354ac52b88d7ad8e4", - "papers/tables/T14b_specification_diagnostics.csv": "7669deb2887fb69f6a6bbfa0f3affb80055ad72ca55b9da0b2a56c724885e586", - "papers/tables/T14b_specification_diagnostics.md": "8afdc57d37a4e25b9c96881091482464ad93d094c25243505419a0be06d2a3e2", - "papers/tables/T15_threshold_grid.csv": "84601c9a63851da30268ead0789a03d88c930dacd282d5b42c6b1d1ee59b56ca", - "papers/tables/T15_threshold_grid.md": "da2a3e642512b52156b7e3fceb145e3c568b67b1acd9c9702d1686ecc5460ddc", - "papers/tables/T16_uncertainty_budget.csv": "e7b14de72c72916c027b039e5e07394be473f7699445b56866c10931bec7d73a", + "papers/tables/T12_dependence_landscape.csv": "49d0bd6b87df52032edda249539411e8949f0824511639754ce11200dddf05ea", + "papers/tables/T12_dependence_landscape.md": "655ecd1faa6b9930ebd7430f73918b715a8191106c281875166dcbbb527b4024", + "papers/tables/T12b_dependence_landscape_taste_motor.csv": "b603d0d154161032c9ee3a083ca8b5ebd976f44fc460189de099d09600a059b7", + "papers/tables/T12b_dependence_landscape_taste_motor.md": "87069c584f977eb7536c5b2a9af953c1bca0bbda1841aeeeb7aa40efe4ac181b", + "papers/tables/T13_ablation.csv": "e78aa24f0941ab855a251647e4af0dee36f5147fbaa356ca1036af0185a10d9f", + "papers/tables/T13_ablation.md": "929adc5e46d634a36dc83b999e759826678ca9b0cc7a7c97fbd9b9c938bbf86d", + "papers/tables/T13b_ablation_information.csv": "666082d138a6b1229b763ac01dad1b8bdbd41a5ef53b907bc8dbe750fed991d0", + "papers/tables/T13b_ablation_information.md": "e7a65038148ced36a75c2282b6309420e7a7860eac87cf4690379192d1ed645d", + "papers/tables/T14_conclusion_stability.csv": "dfec0ed016c1b2e83cc03df683be59da60c7f3ebe0e670865b4a6ec8c424f3c2", + "papers/tables/T14_conclusion_stability.md": "c3a8de7763292ea7f4144695c76be4fca0e9516da26e088eeedcb5d83202f64d", + "papers/tables/T14b_specification_diagnostics.csv": "87b9af5f4edb46fca79724a4225e903aa847156828a48698f525d463696bff8f", + "papers/tables/T14b_specification_diagnostics.md": "977d1dc4c73926e35734b7fe7b2a91504252a9464dc8a908163429c43bd91e52", + "papers/tables/T15_threshold_grid.csv": "ae514a5833030341f3064f29e0b1c305e65afbeb9b10921301c37a0714ddbee9", + "papers/tables/T15_threshold_grid.md": "8d3ba1440e6c2e362825491b0dc609d26930d7a44753a896c150eb9e679e5905", + "papers/tables/T16_uncertainty_budget.csv": "32f5fd5c82ffef36a41075a19e1c61ed63bc7dfcc56b616441b57281318ab7c2", "papers/tables/T16_uncertainty_budget.md": "cdff58f5e9127ca4bcaf8cb267980aa7f13643254411ae4cea134fffb80ce2e8", - "papers/tables/T17_value_of_information.csv": "0383106c6b196411536ed5f4dc4844f0052e23d299fd40db5e59b5c3d1052e63", - "papers/tables/T17_value_of_information.md": "0e4aa934ddf448ede83535f3b8944a0f2bd0e002efd73368fb5337ce59ea4daf", - "papers/tables/T18_claim_provenance.csv": "a9ce64585a892ae6917bd92a4863f2c8d13d170d8e350cabaead19089c144587", - "papers/tables/T18_claim_provenance.md": "c6ffe932b0fea59efea6dc5668f692b73221a19d4f1e385a4baffbeaf167b043", + "papers/tables/T17_value_of_information.csv": "18fc532cd3a2bf62e990c3f6394abba04a8772a36de40e2f38636d70e00eec87", + "papers/tables/T17_value_of_information.md": "08efd03ef2fa07cedafc8e9201113c46fcd9ba8243f1b37464261a204c4751e4", + "papers/tables/T18_claim_provenance.csv": "79e56df07cbc8ab57062210e1d81b20ad38e6b0a18fd6bf540eb8cc1376976fa", + "papers/tables/T18_claim_provenance.md": "2bf4b8b73af511f0b298b345b58d14a51bd2a9c8038b32518a8278a4a7a3e233", "papers/tables/T19_cut_census.csv": "baa387d8621ed7cc0978e707d03ec304a2b92bb78fdbc950dcc9638b819e438b", "papers/tables/T19_cut_census.md": "32c6337750879dda48123f848b795cd84b5f4ba1a0353f786c069a5dfbc2461b", - "papers/tables/T1_library.csv": "e910be5808eabb7d73ed06583722c16619ad15abed4aee351a7909e22b94f8ea", - "papers/tables/T1_library.md": "a2f5ca4b388ebf6d07f2845fa43e5ed1a1f3f49c3ac0056d1bf0421fefa7436c", + "papers/tables/T1_library.csv": "95e2358c191a6679deae8681eae7ad47aa21941c37caee452edd02b889b4b20c", + "papers/tables/T1_library.md": "1470a15dc41b0a2f74345c0914e8c1526edd2e100663250237b376dc28210e67", "papers/tables/T20_transformation_rule.csv": "204aa33dd9d26703768025fb07bb98afa8ded052683110e03fd30fdcb4bcaf53", "papers/tables/T20_transformation_rule.md": "5b47011470411ee6647ae181f2c6ffe44d4c9967f117a7bf5e38d6180c75cadb", "papers/tables/T21_transmitter_signs.csv": "a116af54def25de5e2bb86a0b133460aee2b2e15db7d424115cd3d8d7b9c133b", "papers/tables/T21_transmitter_signs.md": "92f018dd28843692f00d08037192930b53ee2cc524597241120be914ff5cedb8", "papers/tables/T22_instrument_validation.csv": "f8c3949515d6c250e6edc3f0a6f46ffcdff8a77b9d378d87a02d7948ef4a302c", - "papers/tables/T22_instrument_validation.md": "3c6ea0d076897d92cbbf97878234fa92f2f6f92ba48bcec756fe15144f0c8f3f", + "papers/tables/T22_instrument_validation.md": "0ed7b7f4dcea39de37324f8eb6891cb919eb0a5de25946ec294dcd909f8176c7", "papers/tables/T23_transmitter_null_balance.csv": "e4637660da0612c15fe5ab4138d757c1382a6b4bc697a530e584711b343ba8e6", "papers/tables/T23_transmitter_null_balance.md": "190f7af5782f8169875f3fef5f151db15352a3a0f518d1ab690eb83b0a4042e9", - "papers/tables/T24_composition_reference.csv": "cc66a68aa48ef82c1597fed6b257dd4ebf123f85fd14a785448d07bb066bdd9d", - "papers/tables/T24_composition_reference.md": "17000d72ab57f8ccb489260277782c8191a31d1497f33413fb53d63830082182", - "papers/tables/T25_engine_normalisation.csv": "e04138afc1b6f04c37ef886fcfebb223aec014733dfaae69aec99311f0731951", - "papers/tables/T25_engine_normalisation.md": "1cf6d053d13e7cac9382b206bed3b8379e0c836694faa669b257c832be847d5d", + "papers/tables/T24_composition_reference.csv": "11e2db02ccef3f20563f1575136da03a776b35de05b62c9787482e5d67e608c5", + "papers/tables/T24_composition_reference.md": "966ab19bbf1b1e915cd802c8e4d54519159738a8129cbdb5aee82b80ade879a0", + "papers/tables/T25_engine_normalisation.csv": "f266061cbebbe6ce4de4fb21933771c5584880b3c3c4d1247dabce4a2197c6fd", + "papers/tables/T25_engine_normalisation.md": "40b8bf9d96c49347147e884519edc06f305e217a4a6fe3504f868249cb6afa0f", "papers/tables/T26_scale_frontier.csv": "f2a497c778e69ff6906f263bf664c18be1bf9657e8d339b94844e4d6f0bee52a", "papers/tables/T26_scale_frontier.md": "c942dc98d36c845b22f9da433e1f6d3dfa1d1925c7576118e3473ae15cf3a594", "papers/tables/T27_scale_study.csv": "71493be2c495cbebe866a187f0a380dbc1e9c9b2e4f00de3983bd48290b78ad5", "papers/tables/T27_scale_study.md": "9a01845e385d98af33de3073123dc5422cdd97779bf773ec67cf526ef921ab0a", "papers/tables/T2_mechanisms.csv": "0a7bb3009746100d51946e590d5a7eb23a628e67ea4d99967c3504a1f3deed59", "papers/tables/T2_mechanisms.md": "01adf76048f99904933dae79fb8a5e4e41625c477cf344eff889db73982ba102", - "papers/tables/T3_predictions.csv": "07c2b18ae2d0dfc01fbc74046ce80db268a4fa607eb86660e5de180402a63dcb", - "papers/tables/T3_predictions.md": "d041e7783a5a2bcd52a01f05003e374e9ff7f16082f1f65ac483a39129bc1214", + "papers/tables/T3_predictions.csv": "ac1bb32219f0efb985bb99af5b6da46f54f731d2f3daf53bf055966efcdc8bdd", + "papers/tables/T3_predictions.md": "5e4714f00a873e4697b22d5897a21c808fa2e64233b319a5aaf062a21525ad3b", "papers/tables/T4_rank_validation.csv": "b2cea8431687fb45340c08cefa41364c305e662125cd346ace710314a13ba260", "papers/tables/T4_rank_validation.md": "c90eb1e87097471b37139f676b9a9d9f762e3159200d4179ba2b28b9a41951de", - "papers/tables/T5_mixtures.csv": "89faa9e76bc1313d27aa1369973ef404362c0f52f34830521bd71b3dd9ca91a6", + "papers/tables/T5_mixtures.csv": "112d8e7cb7673adec3d46799501ed4ae6b5fc7372526d81a7020cfd983d56367", "papers/tables/T5_mixtures.md": "3e6f65096e6f9b07b686089f4f76c4b332191deeef846197cdd0a707775bc3d8", "papers/tables/T6_dependence_profile.csv": "6fefaee56cfb66c84e9128b3a198da4f2236dcd9ad01495f7518c605f31c93be", - "papers/tables/T6_dependence_profile.md": "fd383f7e44430e3abde00e7601df427c90a1cf81df6e42c7aeae152b1b4febe0", + "papers/tables/T6_dependence_profile.md": "25e25b3f2f07e23074601f2d6b711f68ba74325b56efc9a3c788ef058cdb540f", "papers/tables/T7_top_paths.csv": "42c9802e0901a9ec55b5c86eb0c5aa81d548dc2807c7ad67c4d73bb50e1dafd9", "papers/tables/T7_top_paths.md": "898e2c0fc95b89c9c57b75fe8338c9a80248a0915febdc650c3058a86596aa52", - "papers/tables/T8_landscape.csv": "415ec69baf7cef49aa8bedfb87aa46c503e3e6fb5f16464bf82d46756ada5c1b", - "papers/tables/T8_landscape.md": "fdf7e86588961bd83728b6bf57f7cfc089fcf1e49197e3e5c76473ef74d250f1", + "papers/tables/T8_landscape.csv": "f786469d72b51eaf41989498aefd1b818ad99ea2a270724f5203ba4155508e36", + "papers/tables/T8_landscape.md": "44da6eec4064822a5dff9c1d8bb3f95f0902645ffe62e949b187bfe231a778e6", "papers/tables/T9_genotype.csv": "b41a67232c92b459320e340b8131faf43e99c186349e94a23ccbd6bc6dbf9a5b", "papers/tables/T9_genotype.md": "7ec3afa5fc5982ccd381da46928a08911129401f4c745e999e092eaec0ff35c1" }, @@ -305,8 +311,8 @@ "papers/figures/*", "papers/tables/*" ], - "sha256": "90db07af3e8418052c470b06c8cdd57bd19fb9081b571c7a39090aeed15b0b74" + "sha256": "fb56e4fae1009d65c4074ba2db77921862b3bf8eb19a244914940a0d0db450ab" } }, - "sha256": "011a7800b53b5fde49eae1ae4111245b9d2597dce6a00c97a9da0c4443e180fb" + "sha256": "7d341e3a293eb192f677fbac98421dc60ad2c30a4c9d971cff0ec5ca0c500f71" } diff --git a/data/derived/README.md b/data/derived/README.md index 6739c6f..1c45ccf 100644 --- a/data/derived/README.md +++ b/data/derived/README.md @@ -25,3 +25,14 @@ changing underneath it. **Only the 1k rung is small enough for git** (1.1 MB); `malecns-scale-ladder` artifact and are not committed. Drop them anywhere on `flylab.circuit.rate.DATA_DIRS` (e.g. `~/.flylab/`) and `flylab.analysis.scale.available_cuts()` will find them. + +`scale_study_high_power_2026-09-24.json` is a separate exploratory rerun of +imidacloprid and fipronil on the 5k–50k rungs. It records all eight profiles, +the four rebuilt-cut hashes, the three public input Feather hashes, code hashes, +and per-profile runtimes. It does not replace the original `papers/scale_study.json` +or its T27 results: the historical larger-cut hashes were never recorded, and +these follow-up profiles use n = 50 or 100 with no across-profile multiplicity +adjustment. Its `path` values identify the temporary files used on the run host; +the 5k–50k graph files are not distributed in this checkout. Rebuild them from +the public Feather inputs with `flylab.maps.extract.extract_ladder` and compare +the resulting hashes with the report before rerunning a profile. diff --git a/data/derived/scale_study_high_power_2026-09-24.json b/data/derived/scale_study_high_power_2026-09-24.json new file mode 100644 index 0000000..1c4e7a8 --- /dev/null +++ b/data/derived/scale_study_high_power_2026-09-24.json @@ -0,0 +1,2616 @@ +{ + "schema_version": 1, + "study": "higher-permutation-power scale follow-up", + "created_at_utc": "2026-09-24T15:13:44.879531+00:00", + "source_map": "MaleCNS v1.0", + "source_url": "https://storage.googleapis.com/flyem-male-cns/v1.0/connectome-data/flat-connectome/", + "source_files_sha256": { + "body-annotations-male-cns-v1.0-minconf-0.5.feather": "2177e246113e4cfbf1e7772ec37c6da1955ff22e8063d0b1f833101f99a9a3b2", + "body-neurotransmitters-male-cns-v1.0.feather": "95c9289220663abeb3409f3ad9e5a7f8a53f8093f5139d15502cd08da8879621", + "connectome-weights-male-cns-v1.0-minconf-0.5.feather": "e35da783d1c686b2b58b3b87cd6a403ae43bfcfba8bff28e08ef752c1a56afc1" + }, + "analysis_code_sha256": { + "flylab/analysis/dependence.py": "b20afd704c5a94c3500965e34f21dc33f0c72e833e24553566190273cd029047", + "flylab/analysis/nullmodels.py": "5f93a5a47945ebe412077d006decc8cc538ca3c2bccb55e48609377a08be48e0", + "flylab/analysis/scale.py": "6264d2ce3497b1cbc075617b3d2ec4f532e092e5b8646aa665cdd76d6615f581", + "flylab/circuit/rate.py": "3f47664856d6fa56146dfab6c900082ea847ce81cad8d88f5317c47f2e2dd3f3" + }, + "parameters": { + "concentration_M": 1e-06, + "seed": 0, + "alpha": 0.05, + "n_jobs": 4, + "permutations_by_cut": { + "scale_5k": 100, + "scale_10k": 50, + "scale_25k": 50, + "scale_50k": 50 + }, + "compounds": [ + "imidacloprid", + "fipronil" + ], + "interpretation": "exploratory; every profile uses fewer than 1000 permutations" + }, + "cuts": { + "scale_5k": { + "path": "/private/tmp/flylab-ladder/malecns_scale_5k.json", + "sha256": "89807461d19674cb784acfcdab3ff130e73819a8e2111ed947a8cbbd33e07850", + "n_nodes": 5000, + "n_edges": 279845 + }, + "scale_10k": { + "path": "/private/tmp/flylab-ladder/malecns_scale_10k.json", + "sha256": "4693b248a8e0e0962027a5c55e946c2462e5e8901d643d01063101ba32a99047", + "n_nodes": 10000, + "n_edges": 621600 + }, + "scale_25k": { + "path": "/private/tmp/flylab-ladder/malecns_scale_25k.json", + "sha256": "cccd19c37188ba5cfb420987993c9ffc723008d5cb398d12d599e70284d6b0d3", + "n_nodes": 25000, + "n_edges": 1364375 + }, + "scale_50k": { + "path": "/private/tmp/flylab-ladder/malecns_scale_50k.json", + "sha256": "ec9c867c1da7d89c52ad2de3a0ebe39d6b1638b536cdeded4824d3bc24aaa48a", + "n_nodes": 50000, + "n_edges": 2216881 + } + }, + "profiles": [ + { + "cut": "scale_5k", + "cut_sha256": "89807461d19674cb784acfcdab3ff130e73819a8e2111ed947a8cbbd33e07850", + "compound": "imidacloprid", + "n": 100, + "wall_seconds": 457.17045962496195, + "profile": { + "compound": "imidacloprid", + "conc_M": 1e-06, + "assay": "subgraph", + "graph": "/private/tmp/flylab-ladder/malecns_scale_5k.json", + "readout": "mean_hz", + "n": 100, + "seed": 0, + "alpha": 0.05, + "effect_floor": 0.01304315074554684, + "effect_floor_absolute": 1e-06, + "effect_floor_frac": 0.01, + "effect_floor_relative": 0.01304315074554684, + "effect_floor_binding": "relative", + "real_effect": -0.6780034084602358, + "real_vehicle": 1.3043150745546839, + "delta": 0.0652157537277342, + "delta_frac": 0.05, + "delta_scale": "0.05 x |vehicle readout| = 0.0652158", + "confirmatory": false, + "multiplicity_correction": null, + "design": "exploratory", + "p": { + "p_sign": 0.0297029702970297, + "p_weight": 0.009900990099009901, + "p_degree": 0.009900990099009901, + "p_ER": 0.009900990099009901, + "p_sign_wm": 0.019801980198019802 + }, + "q": { + "q_sign": null, + "q_weight": null, + "q_degree": null, + "q_ER": null, + "q_sign_wm": null + }, + "p_resolution": 0.009900990099009901, + "resolution_coarser_than_alpha": false, + "D": { + "z_sign": -2.505517847033775, + "z_weight": 3.6575880010104065, + "z_degree": -27.21737410366016, + "z_ER": 8.002686370299607, + "z_sign_wm": -2.451394735094116 + }, + "D_vector": [ + -2.505517847033775, + 3.6575880010104065, + -27.21737410366016, + 8.002686370299607 + ], + "D_keys": [ + "z_sign", + "z_weight", + "z_degree", + "z_ER" + ], + "modes_run": [ + "sign_permute", + "weight_permute", + "rewire_degree_preserving", + "erdos_renyi", + "sign_permute_weight_matched" + ], + "modes": [ + { + "mode": "sign_permute", + "information_kept": "the real edges, the real weights and the unweighted NT histogram", + "rank": 3, + "real_effect": -0.6780034084602358, + "p_two_sided": 0.0297029702970297, + "p_resolution": 0.009900990099009901, + "beats_null": true, + "stabilised": false, + "n_stabilised": null, + "resolution_limited": false, + "verdict_stable": false, + "z": -2.505517847033775, + "null_mean": -0.37859585562152853, + "null_median": -0.35497219645209444, + "null_sd": 0.11949926965923191, + "n": 100, + "n_ok": 100, + "runtime_s": 9.7456119999988, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.0297 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.32303121200814133, + "delta": 0.0652157537277342, + "p_used": 0.0297029702970297, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "weight_permute", + "information_kept": "the real edge list and the transmitters", + "rank": 2, + "real_effect": -0.6780034084602358, + "p_two_sided": 0.009900990099009901, + "p_resolution": 0.009900990099009901, + "beats_null": true, + "stabilised": false, + "n_stabilised": 50, + "resolution_limited": true, + "verdict_stable": true, + "z": 3.6575880010104065, + "null_mean": -1.1020244498347938, + "null_median": -1.090500790793854, + "null_sd": 0.11592914271848617, + "n": 100, + "n_ok": 100, + "runtime_s": 9.040055458026472, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.009901 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.4124973823336182, + "delta": 0.0652157537277342, + "p_used": 0.009900990099009901, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "rewire_degree_preserving", + "information_kept": "every node's in/out degree and its transmitter", + "rank": 1, + "real_effect": -0.6780034084602358, + "p_two_sided": 0.009900990099009901, + "p_resolution": 0.009900990099009901, + "beats_null": true, + "stabilised": false, + "n_stabilised": 50, + "resolution_limited": true, + "verdict_stable": true, + "z": -27.21737410366016, + "null_mean": -0.267058254535597, + "null_median": -0.2647785128068262, + "null_sd": 0.015098633408186698, + "n": 100, + "n_ok": 100, + "runtime_s": 401.70674141601194, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.009901 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.4132248956534096, + "delta": 0.0652157537277342, + "p_used": 0.009900990099009901, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "erdos_renyi", + "information_kept": "N, E, the weight histogram and the transmitter census", + "rank": 0, + "real_effect": -0.6780034084602358, + "p_two_sided": 0.009900990099009901, + "p_resolution": 0.009900990099009901, + "beats_null": true, + "stabilised": false, + "n_stabilised": 50, + "resolution_limited": true, + "verdict_stable": true, + "z": 8.002686370299607, + "null_mean": -0.9739966768991037, + "null_median": -0.9761085747810914, + "null_sd": 0.036986738545370046, + "n": 100, + "n_ok": 100, + "runtime_s": 19.817683333007153, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.009901 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.2981051663208556, + "delta": 0.0652157537277342, + "p_used": 0.009900990099009901, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "sign_permute_weight_matched", + "information_kept": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "rank": 3, + "real_effect": -0.6780034084602358, + "p_two_sided": 0.019801980198019802, + "p_resolution": 0.009900990099009901, + "beats_null": true, + "stabilised": true, + "n_stabilised": 25, + "resolution_limited": false, + "verdict_stable": true, + "z": -2.451394735094116, + "null_mean": -0.41325293131056773, + "null_median": -0.4048203558177801, + "null_sd": 0.1079999370805141, + "n": 100, + "n_ok": 100, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.0198 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.2731830526424557, + "delta": 0.0652157537277342, + "p_used": 0.019801980198019802, + "alpha": 0.05, + "p_basis": "raw" + } + ], + "verdicts": { + "sign_permute": "distinguishable", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "distinguishable" + }, + "classification": { + "class": "topology-dependent", + "reason": "the effect is distinguishable from weight_permute and rewire_degree_preserving, so it needs the wiring pattern and not only the graph's size, degree sequence and transmitter census.", + "alpha": 0.05, + "effect_floor": 0.01304315074554684, + "delta": 0.0652157537277342, + "mode_verdicts": { + "sign_permute": "distinguishable", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "distinguishable" + }, + "mode_verdict_reasons": { + "sign_permute": "the empirical two-sided permutation p = 0.0297 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "weight_permute": "the empirical two-sided permutation p = 0.009901 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "rewire_degree_preserving": "the empirical two-sided permutation p = 0.009901 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "erdos_renyi": "the empirical two-sided permutation p = 0.009901 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "sign_permute_weight_matched": "the empirical two-sided permutation p = 0.0198 is at or below alpha = 0.05: the real graph and this graph model give different drug effects." + }, + "within_tolerance": { + "sign_permute": false, + "weight_permute": false, + "rewire_degree_preserving": false, + "erdos_renyi": false, + "sign_permute_weight_matched": false + }, + "distinguishable": { + "sign_permute": true, + "weight_permute": true, + "rewire_degree_preserving": true, + "erdos_renyi": true, + "sign_permute_weight_matched": true + }, + "structural_distinguishable": [ + "weight_permute", + "rewire_degree_preserving" + ], + "equivalent_modes": [], + "indeterminate_modes": [], + "n_indeterminate": 0, + "er_distinguishable": true, + "equivalence": "no mode was shown equivalent within tolerance", + "network_insensitive": false + }, + "class": "topology-dependent", + "stabilised": false, + "verdict_stable": false, + "n_stabilised": 50, + "runtime_s": 457.1701519580092, + "label": "model_derived", + "warnings": [ + "Null models destroy the MaleCNS wiring on purpose: they are controls, not biology. Nothing here is a measurement from a living fly.", + "The headline statistic is the empirical two-sided permutation p, (k+1)/(n+1); with n shuffles its resolution (and floor) is 1/(n+1), so a p at the floor means 'not resolvable with this n', not 'p = 0'.", + "z is reported second and is not a probability: the null distribution on a degraded graph is usually not normal, and a large |z| can simply mean the shuffles collapsed onto one value.", + "model_derived: these numbers describe this simulation and its teaching EC50 library, not a measured drug effect.", + "The dependence class is a documented label built on the per-mode verdicts; within one cell no correction is applied across the four null modes, and the alpha threshold is a reporting convention.", + "'Not distinguishable' is NOT 'equivalent': only a mode whose gap from the null median is below the prespecified margin delta is reported as equivalent_within_tolerance. Everything else that fails to reject is indeterminate and is bounded by the permutation resolution 1/(n+1).", + "sign_permute keeps the transmitter histogram exactly, so a small |z| means the drug effect follows global E/I balance rather than the identity of the cholinergic cells.", + "shuffles were run on 4 processes; shuffle i is seeded by (seed, i) alone, so the draws are identical to a serial run.", + "the permutation p had not stabilised at this n (it moved by 0.047 between checkpoints, tolerance 0.020); the verdict at alpha also changed between checkpoints. Increase n before quoting the value.", + "weight_permute keeps topology and transmitters, so it isolates the pairing between synapse counts and wiring.", + "p is at its resolution floor 0.0099: with 100 shuffles this is the smallest p obtainable, so report it as 'p <= floor', not as an exact value.", + "the permutation p had not stabilised at this n (it moved by 0.029 between checkpoints, tolerance 0.020); the verdict at alpha never changed, so only the precision of p is at stake. Increase n before quoting the value.", + "rewire_degree_preserving keeps every node's in/out degree and transmitter, so it isolates the wiring pattern itself.", + "erdos_renyi keeps only N, E and the weight histogram; it is the weakest null and should be the easiest to beat.", + "exploratory: this profile is not one of the prespecified confirmatory tests (imidacloprid, fipronil at n >= 1000) and carries no multiplicity correction of its own. Counting verdicts over many such profiles needs the component-test-adjusted dependence_landscape instead." + ], + "necessary_information_level": { + "level": "real_connectome", + "mode": "real_connectome", + "rank": 4, + "keeps": "the MaleCNS cut itself", + "destroys": "nothing", + "p_two_sided": null, + "verdict": "distinguishable", + "equivalent_within_tolerance": false, + "delta": 0.0652157537277342, + "description": "every degradation of the graph was distinguishable from the real one, so this prediction needs the real MaleCNS wiring.", + "class": "topology-dependent", + "warnings": [] + } + } + }, + { + "cut": "scale_5k", + "cut_sha256": "89807461d19674cb784acfcdab3ff130e73819a8e2111ed947a8cbbd33e07850", + "compound": "fipronil", + "n": 100, + "wall_seconds": 445.0752414169838, + "profile": { + "compound": "fipronil", + "conc_M": 1e-06, + "assay": "subgraph", + "graph": "/private/tmp/flylab-ladder/malecns_scale_5k.json", + "readout": "mean_hz", + "n": 100, + "seed": 0, + "alpha": 0.05, + "effect_floor": 0.01304315074554684, + "effect_floor_absolute": 1e-06, + "effect_floor_frac": 0.01, + "effect_floor_relative": 0.01304315074554684, + "effect_floor_binding": "relative", + "real_effect": 0.3361071765645265, + "real_vehicle": 1.3043150745546839, + "delta": 0.0652157537277342, + "delta_frac": 0.05, + "delta_scale": "0.05 x |vehicle readout| = 0.0652158", + "confirmatory": false, + "multiplicity_correction": null, + "design": "exploratory", + "p": { + "p_sign": 0.0891089108910891, + "p_weight": 0.009900990099009901, + "p_degree": 0.009900990099009901, + "p_ER": 0.009900990099009901, + "p_sign_wm": 0.09900990099009901 + }, + "q": { + "q_sign": null, + "q_weight": null, + "q_degree": null, + "q_ER": null, + "q_sign_wm": null + }, + "p_resolution": 0.009900990099009901, + "resolution_coarser_than_alpha": false, + "D": { + "z_sign": -1.7215001409993447, + "z_weight": -5.381520079113035, + "z_degree": 31.60167880342816, + "z_ER": -15.565901448369123, + "z_sign_wm": -1.6247844116058405 + }, + "D_vector": [ + -1.7215001409993447, + -5.381520079113035, + 31.60167880342816, + -15.565901448369123 + ], + "D_keys": [ + "z_sign", + "z_weight", + "z_degree", + "z_ER" + ], + "modes_run": [ + "sign_permute", + "weight_permute", + "rewire_degree_preserving", + "erdos_renyi", + "sign_permute_weight_matched" + ], + "modes": [ + { + "mode": "sign_permute", + "information_kept": "the real edges, the real weights and the unweighted NT histogram", + "rank": 3, + "real_effect": 0.3361071765645265, + "p_two_sided": 0.0891089108910891, + "p_resolution": 0.009900990099009901, + "beats_null": false, + "stabilised": false, + "n_stabilised": 50, + "resolution_limited": false, + "verdict_stable": true, + "z": -1.7215001409993447, + "null_mean": 0.47301612459850645, + "null_median": 0.4673117616209303, + "null_sd": 0.07952886251551697, + "n": 100, + "n_ok": 100, + "runtime_s": 5.394133208959829, + "verdict": "indeterminate", + "verdict_reason": "the test did not reject (p = 0.08911) but the gap between the real effect and the null median, 0.1312, is not below the prespecified margin delta = 0.06522: failure to reject is not evidence of equivalence.", + "distinguishable": false, + "within_tolerance": false, + "abs_gap_from_null_median": 0.1312045850564038, + "delta": 0.0652157537277342, + "p_used": 0.0891089108910891, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "weight_permute", + "information_kept": "the real edge list and the transmitters", + "rank": 2, + "real_effect": 0.3361071765645265, + "p_two_sided": 0.009900990099009901, + "p_resolution": 0.009900990099009901, + "beats_null": true, + "stabilised": false, + "n_stabilised": 50, + "resolution_limited": true, + "verdict_stable": true, + "z": -5.381520079113035, + "null_mean": 0.5789190007278059, + "null_median": 0.5817544076447131, + "null_sd": 0.04511956112654677, + "n": 100, + "n_ok": 100, + "runtime_s": 5.722207500017248, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.009901 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.2456472310801866, + "delta": 0.0652157537277342, + "p_used": 0.009900990099009901, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "rewire_degree_preserving", + "information_kept": "every node's in/out degree and its transmitter", + "rank": 1, + "real_effect": 0.3361071765645265, + "p_two_sided": 0.009900990099009901, + "p_resolution": 0.009900990099009901, + "beats_null": true, + "stabilised": false, + "n_stabilised": 50, + "resolution_limited": true, + "verdict_stable": true, + "z": 31.60167880342816, + "null_mean": 0.1505405246918745, + "null_median": 0.15065051698499204, + "null_sd": 0.005872050438425494, + "n": 100, + "n_ok": 100, + "runtime_s": 398.12229224998737, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.009901 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.18545665957953444, + "delta": 0.0652157537277342, + "p_used": 0.009900990099009901, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "erdos_renyi", + "information_kept": "N, E, the weight histogram and the transmitter census", + "rank": 0, + "real_effect": 0.3361071765645265, + "p_two_sided": 0.009900990099009901, + "p_resolution": 0.009900990099009901, + "beats_null": true, + "stabilised": false, + "n_stabilised": 50, + "resolution_limited": true, + "verdict_stable": true, + "z": -15.565901448369123, + "null_mean": 0.663663884846884, + "null_median": 0.6655077842706429, + "null_sd": 0.021043221259548474, + "n": 100, + "n_ok": 100, + "runtime_s": 19.457524833036587, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.009901 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.32940060770611646, + "delta": 0.0652157537277342, + "p_used": 0.009900990099009901, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "sign_permute_weight_matched", + "information_kept": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "rank": 3, + "real_effect": 0.3361071765645265, + "p_two_sided": 0.09900990099009901, + "p_resolution": 0.009900990099009901, + "beats_null": false, + "stabilised": false, + "n_stabilised": null, + "resolution_limited": false, + "verdict_stable": true, + "z": -1.6247844116058405, + "null_mean": 0.4483355264712141, + "null_median": 0.4536443130073468, + "null_sd": 0.06907276381102634, + "n": 100, + "n_ok": 100, + "verdict": "indeterminate", + "verdict_reason": "the test did not reject (p = 0.09901) but the gap between the real effect and the null median, 0.1175, is not below the prespecified margin delta = 0.06522: failure to reject is not evidence of equivalence.", + "distinguishable": false, + "within_tolerance": false, + "abs_gap_from_null_median": 0.11753713644282032, + "delta": 0.0652157537277342, + "p_used": 0.09900990099009901, + "alpha": 0.05, + "p_basis": "raw" + } + ], + "verdicts": { + "sign_permute": "indeterminate", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "indeterminate" + }, + "classification": { + "class": "topology-dependent", + "reason": "the effect is distinguishable from weight_permute and rewire_degree_preserving, so it needs the wiring pattern and not only the graph's size, degree sequence and transmitter census.", + "alpha": 0.05, + "effect_floor": 0.01304315074554684, + "delta": 0.0652157537277342, + "mode_verdicts": { + "sign_permute": "indeterminate", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "indeterminate" + }, + "mode_verdict_reasons": { + "sign_permute": "the test did not reject (p = 0.08911) but the gap between the real effect and the null median, 0.1312, is not below the prespecified margin delta = 0.06522: failure to reject is not evidence of equivalence.", + "weight_permute": "the empirical two-sided permutation p = 0.009901 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "rewire_degree_preserving": "the empirical two-sided permutation p = 0.009901 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "erdos_renyi": "the empirical two-sided permutation p = 0.009901 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "sign_permute_weight_matched": "the test did not reject (p = 0.09901) but the gap between the real effect and the null median, 0.1175, is not below the prespecified margin delta = 0.06522: failure to reject is not evidence of equivalence." + }, + "within_tolerance": { + "sign_permute": false, + "weight_permute": false, + "rewire_degree_preserving": false, + "erdos_renyi": false, + "sign_permute_weight_matched": false + }, + "distinguishable": { + "sign_permute": false, + "weight_permute": true, + "rewire_degree_preserving": true, + "erdos_renyi": true, + "sign_permute_weight_matched": false + }, + "structural_distinguishable": [ + "weight_permute", + "rewire_degree_preserving" + ], + "equivalent_modes": [], + "indeterminate_modes": [ + "sign_permute", + "sign_permute_weight_matched" + ], + "n_indeterminate": 2, + "er_distinguishable": true, + "equivalence": "no mode was shown equivalent within tolerance", + "network_insensitive": false + }, + "class": "topology-dependent", + "stabilised": false, + "verdict_stable": true, + "n_stabilised": 50, + "runtime_s": 445.07488104200456, + "label": "model_derived", + "warnings": [ + "Null models destroy the MaleCNS wiring on purpose: they are controls, not biology. Nothing here is a measurement from a living fly.", + "The headline statistic is the empirical two-sided permutation p, (k+1)/(n+1); with n shuffles its resolution (and floor) is 1/(n+1), so a p at the floor means 'not resolvable with this n', not 'p = 0'.", + "z is reported second and is not a probability: the null distribution on a degraded graph is usually not normal, and a large |z| can simply mean the shuffles collapsed onto one value.", + "model_derived: these numbers describe this simulation and its teaching EC50 library, not a measured drug effect.", + "The dependence class is a documented label built on the per-mode verdicts; within one cell no correction is applied across the four null modes, and the alpha threshold is a reporting convention.", + "'Not distinguishable' is NOT 'equivalent': only a mode whose gap from the null median is below the prespecified margin delta is reported as equivalent_within_tolerance. Everything else that fails to reject is indeterminate and is bounded by the permutation resolution 1/(n+1).", + "sign_permute keeps the transmitter histogram exactly, so a small |z| means the drug effect follows global E/I balance rather than the identity of the cholinergic cells.", + "shuffles were run on 4 processes; shuffle i is seeded by (seed, i) alone, so the draws are identical to a serial run.", + "the permutation p had not stabilised at this n (it moved by 0.103 between checkpoints, tolerance 0.020); the verdict at alpha never changed, so only the precision of p is at stake. Increase n before quoting the value.", + "weight_permute keeps topology and transmitters, so it isolates the pairing between synapse counts and wiring.", + "p is at its resolution floor 0.0099: with 100 shuffles this is the smallest p obtainable, so report it as 'p <= floor', not as an exact value.", + "the permutation p had not stabilised at this n (it moved by 0.029 between checkpoints, tolerance 0.020); the verdict at alpha never changed, so only the precision of p is at stake. Increase n before quoting the value.", + "rewire_degree_preserving keeps every node's in/out degree and transmitter, so it isolates the wiring pattern itself.", + "erdos_renyi keeps only N, E and the weight histogram; it is the weakest null and should be the easiest to beat.", + "indeterminate (neither distinguishable nor equivalent within delta = 0.06522): sign_permute, sign_permute_weight_matched. These modes support 'not distinguishable from this null ensemble' and nothing stronger.", + "exploratory: this profile is not one of the prespecified confirmatory tests (imidacloprid, fipronil at n >= 1000) and carries no multiplicity correction of its own. Counting verdicts over many such profiles needs the component-test-adjusted dependence_landscape instead.", + "sign_permute_weight_matched: the permutation p had not stabilised at n=100, so this level is provisional.", + "sign_permute_weight_matched is INDETERMINATE, not equivalent: the necessary information level is bounded by what n shuffles could resolve. Report it as 'not distinguishable from the real graph', never as 'this graph model gives the same effect'." + ], + "necessary_information_level": { + "level": "wiring_with_weighted_transmitter_balance", + "mode": "sign_permute_weight_matched", + "rank": 3, + "keeps": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "destroys": "which cell carries which transmitter, at matched out-strength", + "p_two_sided": 0.09900990099009901, + "p_used": 0.09900990099009901, + "p_basis": "raw", + "verdict": "indeterminate", + "equivalent_within_tolerance": false, + "abs_gap_from_null_median": 0.11753713644282032, + "delta": 0.0652157537277342, + "non_monotone": false, + "richer_models_distinguishable": [], + "ladder_substitutions": [], + "description": "a graph model that keeps only the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter could not be distinguished from the real cut on this readout, but the gap (0.1175) is not below the prespecified margin delta = 0.06522, so this is the weakest level this study can rule out needing, not a demonstration that the level suffices.", + "ladder_note": "ranks 2 and 3 are not strictly nested: weight_permute keeps transmitter identity and loses the weight-topology pairing, sign_permute the reverse.", + "class": "topology-dependent", + "warnings": [ + "sign_permute_weight_matched: the permutation p had not stabilised at n=100, so this level is provisional.", + "sign_permute_weight_matched is INDETERMINATE, not equivalent: the necessary information level is bounded by what n shuffles could resolve. Report it as 'not distinguishable from the real graph', never as 'this graph model gives the same effect'." + ] + } + } + }, + { + "cut": "scale_10k", + "cut_sha256": "4693b248a8e0e0962027a5c55e946c2462e5e8901d643d01063101ba32a99047", + "compound": "imidacloprid", + "n": 50, + "wall_seconds": 519.6635590830119, + "profile": { + "compound": "imidacloprid", + "conc_M": 1e-06, + "assay": "subgraph", + "graph": "/private/tmp/flylab-ladder/malecns_scale_10k.json", + "readout": "mean_hz", + "n": 50, + "seed": 0, + "alpha": 0.05, + "effect_floor": 0.006558813707747225, + "effect_floor_absolute": 1e-06, + "effect_floor_frac": 0.01, + "effect_floor_relative": 0.006558813707747225, + "effect_floor_binding": "relative", + "real_effect": -0.34321529468796597, + "real_vehicle": 0.6558813707747224, + "delta": 0.03279406853873612, + "delta_frac": 0.05, + "delta_scale": "0.05 x |vehicle readout| = 0.0327941", + "confirmatory": false, + "multiplicity_correction": null, + "design": "exploratory", + "p": { + "p_sign": 0.0392156862745098, + "p_weight": 0.0196078431372549, + "p_degree": 0.0196078431372549, + "p_ER": 0.0196078431372549, + "p_sign_wm": 0.0196078431372549 + }, + "q": { + "q_sign": null, + "q_weight": null, + "q_degree": null, + "q_ER": null, + "q_sign_wm": null + }, + "p_resolution": 0.0196078431372549, + "resolution_coarser_than_alpha": false, + "D": { + "z_sign": -2.4186546311254697, + "z_weight": 3.5363085691154126, + "z_degree": -32.658762385664055, + "z_ER": 9.534694545269831, + "z_sign_wm": -3.4682674266604687 + }, + "D_vector": [ + -2.4186546311254697, + 3.5363085691154126, + -32.658762385664055, + 9.534694545269831 + ], + "D_keys": [ + "z_sign", + "z_weight", + "z_degree", + "z_ER" + ], + "modes_run": [ + "sign_permute", + "weight_permute", + "rewire_degree_preserving", + "erdos_renyi", + "sign_permute_weight_matched" + ], + "modes": [ + { + "mode": "sign_permute", + "information_kept": "the real edges, the real weights and the unweighted NT histogram", + "rank": 3, + "real_effect": -0.34321529468796597, + "p_two_sided": 0.0392156862745098, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": null, + "resolution_limited": false, + "verdict_stable": false, + "z": -2.4186546311254697, + "null_mean": -0.1970030427779429, + "null_median": -0.19330685475129533, + "null_sd": 0.06045189339082541, + "n": 50, + "n_ok": 50, + "runtime_s": 13.111495374992955, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.03922 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.14990843993667063, + "delta": 0.03279406853873612, + "p_used": 0.0392156862745098, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "weight_permute", + "information_kept": "the real edge list and the transmitters", + "rank": 2, + "real_effect": -0.34321529468796597, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": 3.5363085691154126, + "null_mean": -0.5459235315676079, + "null_median": -0.535288045938157, + "null_sd": 0.057321987863278634, + "n": 50, + "n_ok": 50, + "runtime_s": 13.043119959009346, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.19207275125019108, + "delta": 0.03279406853873612, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "rewire_degree_preserving", + "information_kept": "every node's in/out degree and its transmitter", + "rank": 1, + "real_effect": -0.34321529468796597, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": -32.658762385664055, + "null_mean": -0.13942223239020365, + "null_median": -0.13954082256226563, + "null_sd": 0.006240073028217985, + "n": 50, + "n_ok": 50, + "runtime_s": 452.6182830419857, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.20367447212570033, + "delta": 0.03279406853873612, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "erdos_renyi", + "information_kept": "N, E, the weight histogram and the transmitter census", + "rank": 0, + "real_effect": -0.34321529468796597, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": 9.534694545269831, + "null_mean": -0.5119580155308608, + "null_median": -0.5105629216824362, + "null_sd": 0.01769775843806221, + "n": 50, + "n_ok": 50, + "runtime_s": 22.662949916964862, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.16734762699447026, + "delta": 0.03279406853873612, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "sign_permute_weight_matched", + "information_kept": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "rank": 3, + "real_effect": -0.34321529468796597, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": -3.4682674266604687, + "null_mean": -0.19394992757503687, + "null_median": -0.19258549384137694, + "null_sd": 0.04303744456541345, + "n": 50, + "n_ok": 50, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.15062980084658903, + "delta": 0.03279406853873612, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + } + ], + "verdicts": { + "sign_permute": "distinguishable", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "distinguishable" + }, + "classification": { + "class": "topology-dependent", + "reason": "the effect is distinguishable from weight_permute and rewire_degree_preserving, so it needs the wiring pattern and not only the graph's size, degree sequence and transmitter census.", + "alpha": 0.05, + "effect_floor": 0.006558813707747225, + "delta": 0.03279406853873612, + "mode_verdicts": { + "sign_permute": "distinguishable", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "distinguishable" + }, + "mode_verdict_reasons": { + "sign_permute": "the empirical two-sided permutation p = 0.03922 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "weight_permute": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "rewire_degree_preserving": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "erdos_renyi": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "sign_permute_weight_matched": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects." + }, + "within_tolerance": { + "sign_permute": false, + "weight_permute": false, + "rewire_degree_preserving": false, + "erdos_renyi": false, + "sign_permute_weight_matched": false + }, + "distinguishable": { + "sign_permute": true, + "weight_permute": true, + "rewire_degree_preserving": true, + "erdos_renyi": true, + "sign_permute_weight_matched": true + }, + "structural_distinguishable": [ + "weight_permute", + "rewire_degree_preserving" + ], + "equivalent_modes": [], + "indeterminate_modes": [], + "n_indeterminate": 0, + "er_distinguishable": true, + "equivalence": "no mode was shown equivalent within tolerance", + "network_insensitive": false + }, + "class": "topology-dependent", + "stabilised": false, + "verdict_stable": false, + "n_stabilised": 25, + "runtime_s": 519.6633168749977, + "label": "model_derived", + "warnings": [ + "Null models destroy the MaleCNS wiring on purpose: they are controls, not biology. Nothing here is a measurement from a living fly.", + "The headline statistic is the empirical two-sided permutation p, (k+1)/(n+1); with n shuffles its resolution (and floor) is 1/(n+1), so a p at the floor means 'not resolvable with this n', not 'p = 0'.", + "z is reported second and is not a probability: the null distribution on a degraded graph is usually not normal, and a large |z| can simply mean the shuffles collapsed onto one value.", + "model_derived: these numbers describe this simulation and its teaching EC50 library, not a measured drug effect.", + "The dependence class is a documented label built on the per-mode verdicts; within one cell no correction is applied across the four null modes, and the alpha threshold is a reporting convention.", + "'Not distinguishable' is NOT 'equivalent': only a mode whose gap from the null median is below the prespecified margin delta is reported as equivalent_within_tolerance. Everything else that fails to reject is indeterminate and is bounded by the permutation resolution 1/(n+1).", + "sign_permute keeps the transmitter histogram exactly, so a small |z| means the drug effect follows global E/I balance rather than the identity of the cholinergic cells.", + "shuffles were run on 4 processes; shuffle i is seeded by (seed, i) alone, so the draws are identical to a serial run.", + "the permutation p had not stabilised at this n (it moved by 0.115 between checkpoints, tolerance 0.020); the verdict at alpha also changed between checkpoints. Increase n before quoting the value.", + "weight_permute keeps topology and transmitters, so it isolates the pairing between synapse counts and wiring.", + "p is at its resolution floor 0.0196: with 50 shuffles this is the smallest p obtainable, so report it as 'p <= floor', not as an exact value.", + "the permutation p had not stabilised at this n (it moved by 0.057 between checkpoints, tolerance 0.020); the verdict at alpha also changed between checkpoints. Increase n before quoting the value.", + "rewire_degree_preserving keeps every node's in/out degree and transmitter, so it isolates the wiring pattern itself.", + "erdos_renyi keeps only N, E and the weight histogram; it is the weakest null and should be the easiest to beat.", + "exploratory: this profile is not one of the prespecified confirmatory tests (imidacloprid, fipronil at n >= 1000) and carries no multiplicity correction of its own. Counting verdicts over many such profiles needs the component-test-adjusted dependence_landscape instead." + ], + "necessary_information_level": { + "level": "real_connectome", + "mode": "real_connectome", + "rank": 4, + "keeps": "the MaleCNS cut itself", + "destroys": "nothing", + "p_two_sided": null, + "verdict": "distinguishable", + "equivalent_within_tolerance": false, + "delta": 0.03279406853873612, + "description": "every degradation of the graph was distinguishable from the real one, so this prediction needs the real MaleCNS wiring.", + "class": "topology-dependent", + "warnings": [] + } + } + }, + { + "cut": "scale_10k", + "cut_sha256": "4693b248a8e0e0962027a5c55e946c2462e5e8901d643d01063101ba32a99047", + "compound": "fipronil", + "n": 50, + "wall_seconds": 502.02875991596375, + "profile": { + "compound": "fipronil", + "conc_M": 1e-06, + "assay": "subgraph", + "graph": "/private/tmp/flylab-ladder/malecns_scale_10k.json", + "readout": "mean_hz", + "n": 50, + "seed": 0, + "alpha": 0.05, + "effect_floor": 0.006558813707747225, + "effect_floor_absolute": 1e-06, + "effect_floor_frac": 0.01, + "effect_floor_relative": 0.006558813707747225, + "effect_floor_binding": "relative", + "real_effect": 0.15123231683668337, + "real_vehicle": 0.6558813707747224, + "delta": 0.03279406853873612, + "delta_frac": 0.05, + "delta_scale": "0.05 x |vehicle readout| = 0.0327941", + "confirmatory": false, + "multiplicity_correction": null, + "design": "exploratory", + "p": { + "p_sign": 0.058823529411764705, + "p_weight": 0.0196078431372549, + "p_degree": 0.0196078431372549, + "p_ER": 0.0196078431372549, + "p_sign_wm": 0.0784313725490196 + }, + "q": { + "q_sign": null, + "q_weight": null, + "q_degree": null, + "q_ER": null, + "q_sign_wm": null + }, + "p_resolution": 0.0196078431372549, + "resolution_coarser_than_alpha": false, + "D": { + "z_sign": -1.9715595218496986, + "z_weight": -4.122011308251283, + "z_degree": 31.9399926272723, + "z_ER": -14.631878270859508, + "z_sign_wm": -1.7612765105861108 + }, + "D_vector": [ + -1.9715595218496986, + -4.122011308251283, + 31.9399926272723, + -14.631878270859508 + ], + "D_keys": [ + "z_sign", + "z_weight", + "z_degree", + "z_ER" + ], + "modes_run": [ + "sign_permute", + "weight_permute", + "rewire_degree_preserving", + "erdos_renyi", + "sign_permute_weight_matched" + ], + "modes": [ + { + "mode": "sign_permute", + "information_kept": "the real edges, the real weights and the unweighted NT histogram", + "rank": 3, + "real_effect": 0.15123231683668337, + "p_two_sided": 0.058823529411764705, + "p_resolution": 0.0196078431372549, + "beats_null": false, + "stabilised": false, + "n_stabilised": null, + "resolution_limited": false, + "verdict_stable": true, + "z": -1.9715595218496986, + "null_mean": 0.2256727202481474, + "null_median": 0.2243642001788058, + "null_sd": 0.03775711693534099, + "n": 50, + "n_ok": 50, + "runtime_s": 11.450153875048272, + "verdict": "indeterminate", + "verdict_reason": "the test did not reject (p = 0.05882) but the gap between the real effect and the null median, 0.07313, is not below the prespecified margin delta = 0.03279: failure to reject is not evidence of equivalence.", + "distinguishable": false, + "within_tolerance": false, + "abs_gap_from_null_median": 0.07313188334212242, + "delta": 0.03279406853873612, + "p_used": 0.058823529411764705, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "weight_permute", + "information_kept": "the real edge list and the transmitters", + "rank": 2, + "real_effect": 0.15123231683668337, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": -4.122011308251283, + "null_mean": 0.2463183959276858, + "null_median": 0.24735054582719201, + "null_sd": 0.023067884093540162, + "n": 50, + "n_ok": 50, + "runtime_s": 12.588204999978188, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.09611822899050865, + "delta": 0.03279406853873612, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "rewire_degree_preserving", + "information_kept": "every node's in/out degree and its transmitter", + "rank": 1, + "real_effect": 0.15123231683668337, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": 31.9399926272723, + "null_mean": 0.07485059622632421, + "null_median": 0.07465897645471997, + "null_sd": 0.002391413219837121, + "n": 50, + "n_ok": 50, + "runtime_s": 437.86140004196204, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.0765733403819634, + "delta": 0.03279406853873612, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "erdos_renyi", + "information_kept": "N, E, the weight histogram and the transmitter census", + "rank": 0, + "real_effect": 0.15123231683668337, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": -14.631878270859508, + "null_mean": 0.32399716571874676, + "null_median": 0.32314359301621837, + "null_sd": 0.01180742797909532, + "n": 50, + "n_ok": 50, + "runtime_s": 21.696748000045773, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.171911276179535, + "delta": 0.03279406853873612, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "sign_permute_weight_matched", + "information_kept": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "rank": 3, + "real_effect": 0.15123231683668337, + "p_two_sided": 0.0784313725490196, + "p_resolution": 0.0196078431372549, + "beats_null": false, + "stabilised": false, + "n_stabilised": null, + "resolution_limited": false, + "verdict_stable": true, + "z": -1.7612765105861108, + "null_mean": 0.2150068856065013, + "null_median": 0.21328178530362896, + "null_sd": 0.03620928819892981, + "n": 50, + "n_ok": 50, + "verdict": "indeterminate", + "verdict_reason": "the test did not reject (p = 0.07843) but the gap between the real effect and the null median, 0.06205, is not below the prespecified margin delta = 0.03279: failure to reject is not evidence of equivalence.", + "distinguishable": false, + "within_tolerance": false, + "abs_gap_from_null_median": 0.06204946846694559, + "delta": 0.03279406853873612, + "p_used": 0.0784313725490196, + "alpha": 0.05, + "p_basis": "raw" + } + ], + "verdicts": { + "sign_permute": "indeterminate", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "indeterminate" + }, + "classification": { + "class": "topology-dependent", + "reason": "the effect is distinguishable from weight_permute and rewire_degree_preserving, so it needs the wiring pattern and not only the graph's size, degree sequence and transmitter census.", + "alpha": 0.05, + "effect_floor": 0.006558813707747225, + "delta": 0.03279406853873612, + "mode_verdicts": { + "sign_permute": "indeterminate", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "indeterminate" + }, + "mode_verdict_reasons": { + "sign_permute": "the test did not reject (p = 0.05882) but the gap between the real effect and the null median, 0.07313, is not below the prespecified margin delta = 0.03279: failure to reject is not evidence of equivalence.", + "weight_permute": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "rewire_degree_preserving": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "erdos_renyi": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "sign_permute_weight_matched": "the test did not reject (p = 0.07843) but the gap between the real effect and the null median, 0.06205, is not below the prespecified margin delta = 0.03279: failure to reject is not evidence of equivalence." + }, + "within_tolerance": { + "sign_permute": false, + "weight_permute": false, + "rewire_degree_preserving": false, + "erdos_renyi": false, + "sign_permute_weight_matched": false + }, + "distinguishable": { + "sign_permute": false, + "weight_permute": true, + "rewire_degree_preserving": true, + "erdos_renyi": true, + "sign_permute_weight_matched": false + }, + "structural_distinguishable": [ + "weight_permute", + "rewire_degree_preserving" + ], + "equivalent_modes": [], + "indeterminate_modes": [ + "sign_permute", + "sign_permute_weight_matched" + ], + "n_indeterminate": 2, + "er_distinguishable": true, + "equivalence": "no mode was shown equivalent within tolerance", + "network_insensitive": false + }, + "class": "topology-dependent", + "stabilised": false, + "verdict_stable": false, + "n_stabilised": 25, + "runtime_s": 502.02855237497715, + "label": "model_derived", + "warnings": [ + "Null models destroy the MaleCNS wiring on purpose: they are controls, not biology. Nothing here is a measurement from a living fly.", + "The headline statistic is the empirical two-sided permutation p, (k+1)/(n+1); with n shuffles its resolution (and floor) is 1/(n+1), so a p at the floor means 'not resolvable with this n', not 'p = 0'.", + "z is reported second and is not a probability: the null distribution on a degraded graph is usually not normal, and a large |z| can simply mean the shuffles collapsed onto one value.", + "model_derived: these numbers describe this simulation and its teaching EC50 library, not a measured drug effect.", + "The dependence class is a documented label built on the per-mode verdicts; within one cell no correction is applied across the four null modes, and the alpha threshold is a reporting convention.", + "'Not distinguishable' is NOT 'equivalent': only a mode whose gap from the null median is below the prespecified margin delta is reported as equivalent_within_tolerance. Everything else that fails to reject is indeterminate and is bounded by the permutation resolution 1/(n+1).", + "sign_permute keeps the transmitter histogram exactly, so a small |z| means the drug effect follows global E/I balance rather than the identity of the cholinergic cells.", + "shuffles were run on 4 processes; shuffle i is seeded by (seed, i) alone, so the draws are identical to a serial run.", + "the permutation p had not stabilised at this n (it moved by 0.095 between checkpoints, tolerance 0.020); the verdict at alpha never changed, so only the precision of p is at stake. Increase n before quoting the value.", + "weight_permute keeps topology and transmitters, so it isolates the pairing between synapse counts and wiring.", + "p is at its resolution floor 0.0196: with 50 shuffles this is the smallest p obtainable, so report it as 'p <= floor', not as an exact value.", + "the permutation p had not stabilised at this n (it moved by 0.057 between checkpoints, tolerance 0.020); the verdict at alpha also changed between checkpoints. Increase n before quoting the value.", + "rewire_degree_preserving keeps every node's in/out degree and transmitter, so it isolates the wiring pattern itself.", + "erdos_renyi keeps only N, E and the weight histogram; it is the weakest null and should be the easiest to beat.", + "indeterminate (neither distinguishable nor equivalent within delta = 0.03279): sign_permute, sign_permute_weight_matched. These modes support 'not distinguishable from this null ensemble' and nothing stronger.", + "exploratory: this profile is not one of the prespecified confirmatory tests (imidacloprid, fipronil at n >= 1000) and carries no multiplicity correction of its own. Counting verdicts over many such profiles needs the component-test-adjusted dependence_landscape instead.", + "sign_permute_weight_matched: the permutation p had not stabilised at n=50, so this level is provisional.", + "sign_permute_weight_matched is INDETERMINATE, not equivalent: the necessary information level is bounded by what n shuffles could resolve. Report it as 'not distinguishable from the real graph', never as 'this graph model gives the same effect'." + ], + "necessary_information_level": { + "level": "wiring_with_weighted_transmitter_balance", + "mode": "sign_permute_weight_matched", + "rank": 3, + "keeps": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "destroys": "which cell carries which transmitter, at matched out-strength", + "p_two_sided": 0.0784313725490196, + "p_used": 0.0784313725490196, + "p_basis": "raw", + "verdict": "indeterminate", + "equivalent_within_tolerance": false, + "abs_gap_from_null_median": 0.06204946846694559, + "delta": 0.03279406853873612, + "non_monotone": false, + "richer_models_distinguishable": [], + "ladder_substitutions": [], + "description": "a graph model that keeps only the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter could not be distinguished from the real cut on this readout, but the gap (0.06205) is not below the prespecified margin delta = 0.03279, so this is the weakest level this study can rule out needing, not a demonstration that the level suffices.", + "ladder_note": "ranks 2 and 3 are not strictly nested: weight_permute keeps transmitter identity and loses the weight-topology pairing, sign_permute the reverse.", + "class": "topology-dependent", + "warnings": [ + "sign_permute_weight_matched: the permutation p had not stabilised at n=50, so this level is provisional.", + "sign_permute_weight_matched is INDETERMINATE, not equivalent: the necessary information level is bounded by what n shuffles could resolve. Report it as 'not distinguishable from the real graph', never as 'this graph model gives the same effect'." + ] + } + } + }, + { + "cut": "scale_25k", + "cut_sha256": "cccd19c37188ba5cfb420987993c9ffc723008d5cb398d12d599e70284d6b0d3", + "compound": "imidacloprid", + "n": 50, + "wall_seconds": 1642.244971667009, + "profile": { + "compound": "imidacloprid", + "conc_M": 1e-06, + "assay": "subgraph", + "graph": "/private/tmp/flylab-ladder/malecns_scale_25k.json", + "readout": "mean_hz", + "n": 50, + "seed": 0, + "alpha": 0.05, + "effect_floor": 0.0027480505903201297, + "effect_floor_absolute": 1e-06, + "effect_floor_frac": 0.01, + "effect_floor_relative": 0.0027480505903201297, + "effect_floor_binding": "relative", + "real_effect": -0.1498151852424205, + "real_vehicle": 0.27480505903201297, + "delta": 0.013740252951600649, + "delta_frac": 0.05, + "delta_scale": "0.05 x |vehicle readout| = 0.0137403", + "confirmatory": false, + "multiplicity_correction": null, + "design": "exploratory", + "p": { + "p_sign": 0.29411764705882354, + "p_weight": 0.0196078431372549, + "p_degree": 0.0196078431372549, + "p_ER": 0.0196078431372549, + "p_sign_wm": 0.13725490196078433 + }, + "q": { + "q_sign": null, + "q_weight": null, + "q_degree": null, + "q_ER": null, + "q_sign_wm": null + }, + "p_resolution": 0.0196078431372549, + "resolution_coarser_than_alpha": false, + "D": { + "z_sign": -1.251541270097695, + "z_weight": 3.9492879516187402, + "z_degree": -13.999133474307017, + "z_ER": 10.81331630629886, + "z_sign_wm": -1.324368885813019 + }, + "D_vector": [ + -1.251541270097695, + 3.9492879516187402, + -13.999133474307017, + 10.81331630629886 + ], + "D_keys": [ + "z_sign", + "z_weight", + "z_degree", + "z_ER" + ], + "modes_run": [ + "sign_permute", + "weight_permute", + "rewire_degree_preserving", + "erdos_renyi", + "sign_permute_weight_matched" + ], + "modes": [ + { + "mode": "sign_permute", + "information_kept": "the real edges, the real weights and the unweighted NT histogram", + "rank": 3, + "real_effect": -0.1498151852424205, + "p_two_sided": 0.29411764705882354, + "p_resolution": 0.0196078431372549, + "beats_null": false, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": false, + "verdict_stable": true, + "z": -1.251541270097695, + "null_mean": -0.11342745689527199, + "null_median": -0.11661378632047159, + "null_sd": 0.029074333556981375, + "n": 50, + "n_ok": 50, + "runtime_s": 28.280544500041287, + "verdict": "indeterminate", + "verdict_reason": "the test did not reject (p = 0.2941) but the gap between the real effect and the null median, 0.0332, is not below the prespecified margin delta = 0.01374: failure to reject is not evidence of equivalence.", + "distinguishable": false, + "within_tolerance": false, + "abs_gap_from_null_median": 0.0332013989219489, + "delta": 0.013740252951600649, + "p_used": 0.29411764705882354, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "weight_permute", + "information_kept": "the real edge list and the transmitters", + "rank": 2, + "real_effect": -0.1498151852424205, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": 3.9492879516187402, + "null_mean": -0.23315636406016982, + "null_median": -0.23221431312570556, + "null_sd": 0.02110283672366542, + "n": 50, + "n_ok": 50, + "runtime_s": 27.711132999975234, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.08239912788328507, + "delta": 0.013740252951600649, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "rewire_degree_preserving", + "information_kept": "every node's in/out degree and its transmitter", + "rank": 1, + "real_effect": -0.1498151852424205, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": -13.999133474307017, + "null_mean": -0.07950932412003446, + "null_median": -0.07919697355747934, + "null_sd": 0.0050221580679561525, + "n": 50, + "n_ok": 50, + "runtime_s": 1469.8855237090029, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.07061821168494115, + "delta": 0.013740252951600649, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "erdos_renyi", + "information_kept": "N, E, the weight histogram and the transmitter census", + "rank": 0, + "real_effect": -0.1498151852424205, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": 10.81331630629886, + "null_mean": -0.25066013926316716, + "null_median": -0.24909940355806875, + "null_sd": 0.009325996869434357, + "n": 50, + "n_ok": 50, + "runtime_s": 61.69648674997734, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.09928421831564826, + "delta": 0.013740252951600649, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "sign_permute_weight_matched", + "information_kept": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "rank": 3, + "real_effect": -0.1498151852424205, + "p_two_sided": 0.13725490196078433, + "p_resolution": 0.0196078431372549, + "beats_null": false, + "stabilised": false, + "n_stabilised": null, + "resolution_limited": false, + "verdict_stable": true, + "z": -1.324368885813019, + "null_mean": -0.10790704807855285, + "null_median": -0.09830022088957364, + "null_sd": 0.031643855131903514, + "n": 50, + "n_ok": 50, + "verdict": "indeterminate", + "verdict_reason": "the test did not reject (p = 0.1373) but the gap between the real effect and the null median, 0.05151, is not below the prespecified margin delta = 0.01374: failure to reject is not evidence of equivalence.", + "distinguishable": false, + "within_tolerance": false, + "abs_gap_from_null_median": 0.051514964352846845, + "delta": 0.013740252951600649, + "p_used": 0.13725490196078433, + "alpha": 0.05, + "p_basis": "raw" + } + ], + "verdicts": { + "sign_permute": "indeterminate", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "indeterminate" + }, + "classification": { + "class": "topology-dependent", + "reason": "the effect is distinguishable from weight_permute and rewire_degree_preserving, so it needs the wiring pattern and not only the graph's size, degree sequence and transmitter census.", + "alpha": 0.05, + "effect_floor": 0.0027480505903201297, + "delta": 0.013740252951600649, + "mode_verdicts": { + "sign_permute": "indeterminate", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "indeterminate" + }, + "mode_verdict_reasons": { + "sign_permute": "the test did not reject (p = 0.2941) but the gap between the real effect and the null median, 0.0332, is not below the prespecified margin delta = 0.01374: failure to reject is not evidence of equivalence.", + "weight_permute": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "rewire_degree_preserving": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "erdos_renyi": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "sign_permute_weight_matched": "the test did not reject (p = 0.1373) but the gap between the real effect and the null median, 0.05151, is not below the prespecified margin delta = 0.01374: failure to reject is not evidence of equivalence." + }, + "within_tolerance": { + "sign_permute": false, + "weight_permute": false, + "rewire_degree_preserving": false, + "erdos_renyi": false, + "sign_permute_weight_matched": false + }, + "distinguishable": { + "sign_permute": false, + "weight_permute": true, + "rewire_degree_preserving": true, + "erdos_renyi": true, + "sign_permute_weight_matched": false + }, + "structural_distinguishable": [ + "weight_permute", + "rewire_degree_preserving" + ], + "equivalent_modes": [], + "indeterminate_modes": [ + "sign_permute", + "sign_permute_weight_matched" + ], + "n_indeterminate": 2, + "er_distinguishable": true, + "equivalence": "no mode was shown equivalent within tolerance", + "network_insensitive": false + }, + "class": "topology-dependent", + "stabilised": false, + "verdict_stable": false, + "n_stabilised": 25, + "runtime_s": 1642.244750791986, + "label": "model_derived", + "warnings": [ + "Null models destroy the MaleCNS wiring on purpose: they are controls, not biology. Nothing here is a measurement from a living fly.", + "The headline statistic is the empirical two-sided permutation p, (k+1)/(n+1); with n shuffles its resolution (and floor) is 1/(n+1), so a p at the floor means 'not resolvable with this n', not 'p = 0'.", + "z is reported second and is not a probability: the null distribution on a degraded graph is usually not normal, and a large |z| can simply mean the shuffles collapsed onto one value.", + "model_derived: these numbers describe this simulation and its teaching EC50 library, not a measured drug effect.", + "The dependence class is a documented label built on the per-mode verdicts; within one cell no correction is applied across the four null modes, and the alpha threshold is a reporting convention.", + "'Not distinguishable' is NOT 'equivalent': only a mode whose gap from the null median is below the prespecified margin delta is reported as equivalent_within_tolerance. Everything else that fails to reject is indeterminate and is bounded by the permutation resolution 1/(n+1).", + "sign_permute keeps the transmitter histogram exactly, so a small |z| means the drug effect follows global E/I balance rather than the identity of the cholinergic cells.", + "shuffles were run on 4 processes; shuffle i is seeded by (seed, i) alone, so the draws are identical to a serial run.", + "the permutation p had not stabilised at this n (it moved by 0.167 between checkpoints, tolerance 0.020); the verdict at alpha never changed, so only the precision of p is at stake. Increase n before quoting the value.", + "weight_permute keeps topology and transmitters, so it isolates the pairing between synapse counts and wiring.", + "p is at its resolution floor 0.0196: with 50 shuffles this is the smallest p obtainable, so report it as 'p <= floor', not as an exact value.", + "the permutation p had not stabilised at this n (it moved by 0.057 between checkpoints, tolerance 0.020); the verdict at alpha also changed between checkpoints. Increase n before quoting the value.", + "rewire_degree_preserving keeps every node's in/out degree and transmitter, so it isolates the wiring pattern itself.", + "erdos_renyi keeps only N, E and the weight histogram; it is the weakest null and should be the easiest to beat.", + "indeterminate (neither distinguishable nor equivalent within delta = 0.01374): sign_permute, sign_permute_weight_matched. These modes support 'not distinguishable from this null ensemble' and nothing stronger.", + "exploratory: this profile is not one of the prespecified confirmatory tests (imidacloprid, fipronil at n >= 1000) and carries no multiplicity correction of its own. Counting verdicts over many such profiles needs the component-test-adjusted dependence_landscape instead.", + "sign_permute_weight_matched: the permutation p had not stabilised at n=50, so this level is provisional.", + "sign_permute_weight_matched is INDETERMINATE, not equivalent: the necessary information level is bounded by what n shuffles could resolve. Report it as 'not distinguishable from the real graph', never as 'this graph model gives the same effect'." + ], + "necessary_information_level": { + "level": "wiring_with_weighted_transmitter_balance", + "mode": "sign_permute_weight_matched", + "rank": 3, + "keeps": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "destroys": "which cell carries which transmitter, at matched out-strength", + "p_two_sided": 0.13725490196078433, + "p_used": 0.13725490196078433, + "p_basis": "raw", + "verdict": "indeterminate", + "equivalent_within_tolerance": false, + "abs_gap_from_null_median": 0.051514964352846845, + "delta": 0.013740252951600649, + "non_monotone": false, + "richer_models_distinguishable": [], + "ladder_substitutions": [], + "description": "a graph model that keeps only the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter could not be distinguished from the real cut on this readout, but the gap (0.05151) is not below the prespecified margin delta = 0.01374, so this is the weakest level this study can rule out needing, not a demonstration that the level suffices.", + "ladder_note": "ranks 2 and 3 are not strictly nested: weight_permute keeps transmitter identity and loses the weight-topology pairing, sign_permute the reverse.", + "class": "topology-dependent", + "warnings": [ + "sign_permute_weight_matched: the permutation p had not stabilised at n=50, so this level is provisional.", + "sign_permute_weight_matched is INDETERMINATE, not equivalent: the necessary information level is bounded by what n shuffles could resolve. Report it as 'not distinguishable from the real graph', never as 'this graph model gives the same effect'." + ] + } + } + }, + { + "cut": "scale_25k", + "cut_sha256": "cccd19c37188ba5cfb420987993c9ffc723008d5cb398d12d599e70284d6b0d3", + "compound": "fipronil", + "n": 50, + "wall_seconds": 1496.3287240829668, + "profile": { + "compound": "fipronil", + "conc_M": 1e-06, + "assay": "subgraph", + "graph": "/private/tmp/flylab-ladder/malecns_scale_25k.json", + "readout": "mean_hz", + "n": 50, + "seed": 0, + "alpha": 0.05, + "effect_floor": 0.0027480505903201297, + "effect_floor_absolute": 1e-06, + "effect_floor_frac": 0.01, + "effect_floor_relative": 0.0027480505903201297, + "effect_floor_binding": "relative", + "real_effect": 0.06387880500666127, + "real_vehicle": 0.27480505903201297, + "delta": 0.013740252951600649, + "delta_frac": 0.05, + "delta_scale": "0.05 x |vehicle readout| = 0.0137403", + "confirmatory": false, + "multiplicity_correction": null, + "design": "exploratory", + "p": { + "p_sign": 0.0196078431372549, + "p_weight": 0.0196078431372549, + "p_degree": 0.0196078431372549, + "p_ER": 0.0196078431372549, + "p_sign_wm": 0.0784313725490196 + }, + "q": { + "q_sign": null, + "q_weight": null, + "q_degree": null, + "q_ER": null, + "q_sign_wm": null + }, + "p_resolution": 0.0196078431372549, + "resolution_coarser_than_alpha": false, + "D": { + "z_sign": -2.7212677595967185, + "z_weight": -3.5658513592965835, + "z_degree": 20.61905510626879, + "z_ER": -16.78223501671294, + "z_sign_wm": -2.0559271276820557 + }, + "D_vector": [ + -2.7212677595967185, + -3.5658513592965835, + 20.61905510626879, + -16.78223501671294 + ], + "D_keys": [ + "z_sign", + "z_weight", + "z_degree", + "z_ER" + ], + "modes_run": [ + "sign_permute", + "weight_permute", + "rewire_degree_preserving", + "erdos_renyi", + "sign_permute_weight_matched" + ], + "modes": [ + { + "mode": "sign_permute", + "information_kept": "the real edges, the real weights and the unweighted NT histogram", + "rank": 3, + "real_effect": 0.06387880500666127, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": -2.7212677595967185, + "null_mean": 0.12043779273494935, + "null_median": 0.12082283038544489, + "null_sd": 0.020784058286373815, + "n": 50, + "n_ok": 50, + "runtime_s": 28.081495957972948, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.056944025378783625, + "delta": 0.013740252951600649, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "weight_permute", + "information_kept": "the real edge list and the transmitters", + "rank": 2, + "real_effect": 0.06387880500666127, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": null, + "resolution_limited": true, + "verdict_stable": false, + "z": -3.5658513592965835, + "null_mean": 0.0977906568246904, + "null_median": 0.0979512424051113, + "null_sd": 0.009510169774636581, + "n": 50, + "n_ok": 50, + "runtime_s": 30.84466991701629, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.034072437398450034, + "delta": 0.013740252951600649, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "rewire_degree_preserving", + "information_kept": "every node's in/out degree and its transmitter", + "rank": 1, + "real_effect": 0.06387880500666127, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": 20.61905510626879, + "null_mean": 0.034798502586209624, + "null_median": 0.0348816951455935, + "null_sd": 0.001410360575233653, + "n": 50, + "n_ok": 50, + "runtime_s": 1346.2111157919862, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.028997109861067766, + "delta": 0.013740252951600649, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "erdos_renyi", + "information_kept": "N, E, the weight histogram and the transmitter census", + "rank": 0, + "real_effect": 0.06387880500666127, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": -16.78223501671294, + "null_mean": 0.1389979218168111, + "null_median": 0.138815515798214, + "null_sd": 0.004476109215211257, + "n": 50, + "n_ok": 50, + "runtime_s": 47.920608290995006, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.07493671079155273, + "delta": 0.013740252951600649, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "sign_permute_weight_matched", + "information_kept": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "rank": 3, + "real_effect": 0.06387880500666127, + "p_two_sided": 0.0784313725490196, + "p_resolution": 0.0196078431372549, + "beats_null": false, + "stabilised": false, + "n_stabilised": null, + "resolution_limited": false, + "verdict_stable": true, + "z": -2.0559271276820557, + "null_mean": 0.0969359201335563, + "null_median": 0.09643876459690257, + "null_sd": 0.01607893328600859, + "n": 50, + "n_ok": 50, + "verdict": "indeterminate", + "verdict_reason": "the test did not reject (p = 0.07843) but the gap between the real effect and the null median, 0.03256, is not below the prespecified margin delta = 0.01374: failure to reject is not evidence of equivalence.", + "distinguishable": false, + "within_tolerance": false, + "abs_gap_from_null_median": 0.0325599595902413, + "delta": 0.013740252951600649, + "p_used": 0.0784313725490196, + "alpha": 0.05, + "p_basis": "raw" + } + ], + "verdicts": { + "sign_permute": "distinguishable", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "indeterminate" + }, + "classification": { + "class": "topology-dependent", + "reason": "the effect is distinguishable from weight_permute and rewire_degree_preserving, so it needs the wiring pattern and not only the graph's size, degree sequence and transmitter census.", + "alpha": 0.05, + "effect_floor": 0.0027480505903201297, + "delta": 0.013740252951600649, + "mode_verdicts": { + "sign_permute": "distinguishable", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "indeterminate" + }, + "mode_verdict_reasons": { + "sign_permute": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "weight_permute": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "rewire_degree_preserving": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "erdos_renyi": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "sign_permute_weight_matched": "the test did not reject (p = 0.07843) but the gap between the real effect and the null median, 0.03256, is not below the prespecified margin delta = 0.01374: failure to reject is not evidence of equivalence." + }, + "within_tolerance": { + "sign_permute": false, + "weight_permute": false, + "rewire_degree_preserving": false, + "erdos_renyi": false, + "sign_permute_weight_matched": false + }, + "distinguishable": { + "sign_permute": true, + "weight_permute": true, + "rewire_degree_preserving": true, + "erdos_renyi": true, + "sign_permute_weight_matched": false + }, + "structural_distinguishable": [ + "weight_permute", + "rewire_degree_preserving" + ], + "equivalent_modes": [], + "indeterminate_modes": [ + "sign_permute_weight_matched" + ], + "n_indeterminate": 1, + "er_distinguishable": true, + "equivalence": "no mode was shown equivalent within tolerance", + "network_insensitive": false + }, + "class": "topology-dependent", + "stabilised": false, + "verdict_stable": false, + "n_stabilised": 25, + "runtime_s": 1496.328508333012, + "label": "model_derived", + "warnings": [ + "Null models destroy the MaleCNS wiring on purpose: they are controls, not biology. Nothing here is a measurement from a living fly.", + "The headline statistic is the empirical two-sided permutation p, (k+1)/(n+1); with n shuffles its resolution (and floor) is 1/(n+1), so a p at the floor means 'not resolvable with this n', not 'p = 0'.", + "z is reported second and is not a probability: the null distribution on a degraded graph is usually not normal, and a large |z| can simply mean the shuffles collapsed onto one value.", + "model_derived: these numbers describe this simulation and its teaching EC50 library, not a measured drug effect.", + "The dependence class is a documented label built on the per-mode verdicts; within one cell no correction is applied across the four null modes, and the alpha threshold is a reporting convention.", + "'Not distinguishable' is NOT 'equivalent': only a mode whose gap from the null median is below the prespecified margin delta is reported as equivalent_within_tolerance. Everything else that fails to reject is indeterminate and is bounded by the permutation resolution 1/(n+1).", + "sign_permute keeps the transmitter histogram exactly, so a small |z| means the drug effect follows global E/I balance rather than the identity of the cholinergic cells.", + "shuffles were run on 4 processes; shuffle i is seeded by (seed, i) alone, so the draws are identical to a serial run.", + "p is at its resolution floor 0.0196: with 50 shuffles this is the smallest p obtainable, so report it as 'p <= floor', not as an exact value.", + "the permutation p had not stabilised at this n (it moved by 0.057 between checkpoints, tolerance 0.020); the verdict at alpha also changed between checkpoints. Increase n before quoting the value.", + "weight_permute keeps topology and transmitters, so it isolates the pairing between synapse counts and wiring.", + "rewire_degree_preserving keeps every node's in/out degree and transmitter, so it isolates the wiring pattern itself.", + "erdos_renyi keeps only N, E and the weight histogram; it is the weakest null and should be the easiest to beat.", + "indeterminate (neither distinguishable nor equivalent within delta = 0.01374): sign_permute_weight_matched. These modes support 'not distinguishable from this null ensemble' and nothing stronger.", + "exploratory: this profile is not one of the prespecified confirmatory tests (imidacloprid, fipronil at n >= 1000) and carries no multiplicity correction of its own. Counting verdicts over many such profiles needs the component-test-adjusted dependence_landscape instead.", + "sign_permute_weight_matched: the permutation p had not stabilised at n=50, so this level is provisional.", + "sign_permute_weight_matched is INDETERMINATE, not equivalent: the necessary information level is bounded by what n shuffles could resolve. Report it as 'not distinguishable from the real graph', never as 'this graph model gives the same effect'." + ], + "necessary_information_level": { + "level": "wiring_with_weighted_transmitter_balance", + "mode": "sign_permute_weight_matched", + "rank": 3, + "keeps": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "destroys": "which cell carries which transmitter, at matched out-strength", + "p_two_sided": 0.0784313725490196, + "p_used": 0.0784313725490196, + "p_basis": "raw", + "verdict": "indeterminate", + "equivalent_within_tolerance": false, + "abs_gap_from_null_median": 0.0325599595902413, + "delta": 0.013740252951600649, + "non_monotone": false, + "richer_models_distinguishable": [], + "ladder_substitutions": [], + "description": "a graph model that keeps only the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter could not be distinguished from the real cut on this readout, but the gap (0.03256) is not below the prespecified margin delta = 0.01374, so this is the weakest level this study can rule out needing, not a demonstration that the level suffices.", + "ladder_note": "ranks 2 and 3 are not strictly nested: weight_permute keeps transmitter identity and loses the weight-topology pairing, sign_permute the reverse.", + "class": "topology-dependent", + "warnings": [ + "sign_permute_weight_matched: the permutation p had not stabilised at n=50, so this level is provisional.", + "sign_permute_weight_matched is INDETERMINATE, not equivalent: the necessary information level is bounded by what n shuffles could resolve. Report it as 'not distinguishable from the real graph', never as 'this graph model gives the same effect'." + ] + } + } + }, + { + "cut": "scale_50k", + "cut_sha256": "ec9c867c1da7d89c52ad2de3a0ebe39d6b1638b536cdeded4824d3bc24aaa48a", + "compound": "imidacloprid", + "n": 50, + "wall_seconds": 2485.6113681669813, + "profile": { + "compound": "imidacloprid", + "conc_M": 1e-06, + "assay": "subgraph", + "graph": "/private/tmp/flylab-ladder/malecns_scale_50k.json", + "readout": "mean_hz", + "n": 50, + "seed": 0, + "alpha": 0.05, + "effect_floor": 0.0014003326665626462, + "effect_floor_absolute": 1e-06, + "effect_floor_frac": 0.01, + "effect_floor_relative": 0.0014003326665626462, + "effect_floor_binding": "relative", + "real_effect": -0.07754306127132168, + "real_vehicle": 0.14003326665626462, + "delta": 0.007001663332813231, + "delta_frac": 0.05, + "delta_scale": "0.05 x |vehicle readout| = 0.00700166", + "confirmatory": false, + "multiplicity_correction": null, + "design": "exploratory", + "p": { + "p_sign": 0.8431372549019608, + "p_weight": 0.0196078431372549, + "p_degree": 0.0196078431372549, + "p_ER": 0.0196078431372549, + "p_sign_wm": 0.5294117647058824 + }, + "q": { + "q_sign": null, + "q_weight": null, + "q_degree": null, + "q_ER": null, + "q_sign_wm": null + }, + "p_resolution": 0.0196078431372549, + "resolution_coarser_than_alpha": false, + "D": { + "z_sign": 0.18803780212666857, + "z_weight": 3.9461014823612413, + "z_degree": -5.3747786669299895, + "z_ER": 13.419893565110288, + "z_sign_wm": -0.6599296321412715 + }, + "D_vector": [ + 0.18803780212666857, + 3.9461014823612413, + -5.3747786669299895, + 13.419893565110288 + ], + "D_keys": [ + "z_sign", + "z_weight", + "z_degree", + "z_ER" + ], + "modes_run": [ + "sign_permute", + "weight_permute", + "rewire_degree_preserving", + "erdos_renyi", + "sign_permute_weight_matched" + ], + "modes": [ + { + "mode": "sign_permute", + "information_kept": "the real edges, the real weights and the unweighted NT histogram", + "rank": 3, + "real_effect": -0.07754306127132168, + "p_two_sided": 0.8431372549019608, + "p_resolution": 0.0196078431372549, + "beats_null": false, + "stabilised": false, + "n_stabilised": null, + "resolution_limited": false, + "verdict_stable": true, + "z": 0.18803780212666857, + "null_mean": -0.08244261946010764, + "null_median": -0.0777598374422279, + "null_sd": 0.026056240465337128, + "n": 50, + "n_ok": 50, + "runtime_s": 45.33693162497366, + "verdict": "equivalent_within_tolerance", + "verdict_reason": "the test did not reject (p = 0.8431) and the gap between the real effect and the null median, 0.0002168, is below the prespecified margin delta = 0.007002: this graph model gives the same effect to within the margin.", + "distinguishable": false, + "within_tolerance": true, + "abs_gap_from_null_median": 0.0002167761709062216, + "delta": 0.007001663332813231, + "p_used": 0.8431372549019608, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "weight_permute", + "information_kept": "the real edge list and the transmitters", + "rank": 2, + "real_effect": -0.07754306127132168, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": 3.9461014823612413, + "null_mean": -0.12483795489060881, + "null_median": -0.12316704004446433, + "null_sd": 0.011985219800020736, + "n": 50, + "n_ok": 50, + "runtime_s": 47.24245304201031, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.04562397877314264, + "delta": 0.007001663332813231, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "rewire_degree_preserving", + "information_kept": "every node's in/out degree and its transmitter", + "rank": 1, + "real_effect": -0.07754306127132168, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": -5.3747786669299895, + "null_mean": -0.05889928490410834, + "null_median": -0.059102083036899206, + "null_sd": 0.003468752393828796, + "n": 50, + "n_ok": 50, + "runtime_s": 2250.581340125005, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.018440978234422478, + "delta": 0.007001663332813231, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "erdos_renyi", + "information_kept": "N, E, the weight histogram and the transmitter census", + "rank": 0, + "real_effect": -0.07754306127132168, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": 13.419893565110288, + "null_mean": -0.1523567480997428, + "null_median": -0.15241377008622914, + "null_sd": 0.005574834589070475, + "n": 50, + "n_ok": 50, + "runtime_s": 73.73060025001178, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.07487070881490746, + "delta": 0.007001663332813231, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "sign_permute_weight_matched", + "information_kept": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "rank": 3, + "real_effect": -0.07754306127132168, + "p_two_sided": 0.5294117647058824, + "p_resolution": 0.0196078431372549, + "beats_null": false, + "stabilised": false, + "n_stabilised": null, + "resolution_limited": false, + "verdict_stable": true, + "z": -0.6599296321412715, + "null_mean": -0.0653503936216316, + "null_median": -0.06500580700826905, + "null_sd": 0.01847570870568211, + "n": 50, + "n_ok": 50, + "verdict": "indeterminate", + "verdict_reason": "the test did not reject (p = 0.5294) but the gap between the real effect and the null median, 0.01254, is not below the prespecified margin delta = 0.007002: failure to reject is not evidence of equivalence.", + "distinguishable": false, + "within_tolerance": false, + "abs_gap_from_null_median": 0.012537254263052633, + "delta": 0.007001663332813231, + "p_used": 0.5294117647058824, + "alpha": 0.05, + "p_basis": "raw" + } + ], + "verdicts": { + "sign_permute": "equivalent_within_tolerance", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "indeterminate" + }, + "classification": { + "class": "topology-dependent", + "reason": "the effect is distinguishable from weight_permute and rewire_degree_preserving, so it needs the wiring pattern and not only the graph's size, degree sequence and transmitter census.", + "alpha": 0.05, + "effect_floor": 0.0014003326665626462, + "delta": 0.007001663332813231, + "mode_verdicts": { + "sign_permute": "equivalent_within_tolerance", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "indeterminate" + }, + "mode_verdict_reasons": { + "sign_permute": "the test did not reject (p = 0.8431) and the gap between the real effect and the null median, 0.0002168, is below the prespecified margin delta = 0.007002: this graph model gives the same effect to within the margin.", + "weight_permute": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "rewire_degree_preserving": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "erdos_renyi": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "sign_permute_weight_matched": "the test did not reject (p = 0.5294) but the gap between the real effect and the null median, 0.01254, is not below the prespecified margin delta = 0.007002: failure to reject is not evidence of equivalence." + }, + "within_tolerance": { + "sign_permute": true, + "weight_permute": false, + "rewire_degree_preserving": false, + "erdos_renyi": false, + "sign_permute_weight_matched": false + }, + "distinguishable": { + "sign_permute": false, + "weight_permute": true, + "rewire_degree_preserving": true, + "erdos_renyi": true, + "sign_permute_weight_matched": false + }, + "structural_distinguishable": [ + "weight_permute", + "rewire_degree_preserving" + ], + "equivalent_modes": [ + "sign_permute" + ], + "indeterminate_modes": [ + "sign_permute_weight_matched" + ], + "n_indeterminate": 1, + "er_distinguishable": true, + "equivalence": "equivalence shown within tolerance for sign_permute", + "network_insensitive": false + }, + "class": "topology-dependent", + "stabilised": false, + "verdict_stable": false, + "n_stabilised": 25, + "runtime_s": 2485.6108414590126, + "label": "model_derived", + "warnings": [ + "Null models destroy the MaleCNS wiring on purpose: they are controls, not biology. Nothing here is a measurement from a living fly.", + "The headline statistic is the empirical two-sided permutation p, (k+1)/(n+1); with n shuffles its resolution (and floor) is 1/(n+1), so a p at the floor means 'not resolvable with this n', not 'p = 0'.", + "z is reported second and is not a probability: the null distribution on a degraded graph is usually not normal, and a large |z| can simply mean the shuffles collapsed onto one value.", + "model_derived: these numbers describe this simulation and its teaching EC50 library, not a measured drug effect.", + "The dependence class is a documented label built on the per-mode verdicts; within one cell no correction is applied across the four null modes, and the alpha threshold is a reporting convention.", + "'Not distinguishable' is NOT 'equivalent': only a mode whose gap from the null median is below the prespecified margin delta is reported as equivalent_within_tolerance. Everything else that fails to reject is indeterminate and is bounded by the permutation resolution 1/(n+1).", + "sign_permute keeps the transmitter histogram exactly, so a small |z| means the drug effect follows global E/I balance rather than the identity of the cholinergic cells.", + "shuffles were run on 4 processes; shuffle i is seeded by (seed, i) alone, so the draws are identical to a serial run.", + "the permutation p had not stabilised at this n (it moved by 0.157 between checkpoints, tolerance 0.020); the verdict at alpha never changed, so only the precision of p is at stake. Increase n before quoting the value.", + "weight_permute keeps topology and transmitters, so it isolates the pairing between synapse counts and wiring.", + "p is at its resolution floor 0.0196: with 50 shuffles this is the smallest p obtainable, so report it as 'p <= floor', not as an exact value.", + "the permutation p had not stabilised at this n (it moved by 0.057 between checkpoints, tolerance 0.020); the verdict at alpha also changed between checkpoints. Increase n before quoting the value.", + "rewire_degree_preserving keeps every node's in/out degree and transmitter, so it isolates the wiring pattern itself.", + "erdos_renyi keeps only N, E and the weight histogram; it is the weakest null and should be the easiest to beat.", + "indeterminate (neither distinguishable nor equivalent within delta = 0.007002): sign_permute_weight_matched. These modes support 'not distinguishable from this null ensemble' and nothing stronger.", + "exploratory: this profile is not one of the prespecified confirmatory tests (imidacloprid, fipronil at n >= 1000) and carries no multiplicity correction of its own. Counting verdicts over many such profiles needs the component-test-adjusted dependence_landscape instead.", + "sign_permute_weight_matched: the permutation p had not stabilised at n=50, so this level is provisional.", + "sign_permute_weight_matched is INDETERMINATE, not equivalent: the necessary information level is bounded by what n shuffles could resolve. Report it as 'not distinguishable from the real graph', never as 'this graph model gives the same effect'." + ], + "necessary_information_level": { + "level": "wiring_with_weighted_transmitter_balance", + "mode": "sign_permute_weight_matched", + "rank": 3, + "keeps": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "destroys": "which cell carries which transmitter, at matched out-strength", + "p_two_sided": 0.5294117647058824, + "p_used": 0.5294117647058824, + "p_basis": "raw", + "verdict": "indeterminate", + "equivalent_within_tolerance": false, + "abs_gap_from_null_median": 0.012537254263052633, + "delta": 0.007001663332813231, + "non_monotone": false, + "richer_models_distinguishable": [], + "ladder_substitutions": [], + "description": "a graph model that keeps only the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter could not be distinguished from the real cut on this readout, but the gap (0.01254) is not below the prespecified margin delta = 0.007002, so this is the weakest level this study can rule out needing, not a demonstration that the level suffices.", + "ladder_note": "ranks 2 and 3 are not strictly nested: weight_permute keeps transmitter identity and loses the weight-topology pairing, sign_permute the reverse.", + "class": "topology-dependent", + "warnings": [ + "sign_permute_weight_matched: the permutation p had not stabilised at n=50, so this level is provisional.", + "sign_permute_weight_matched is INDETERMINATE, not equivalent: the necessary information level is bounded by what n shuffles could resolve. Report it as 'not distinguishable from the real graph', never as 'this graph model gives the same effect'." + ] + } + } + }, + { + "cut": "scale_50k", + "cut_sha256": "ec9c867c1da7d89c52ad2de3a0ebe39d6b1638b536cdeded4824d3bc24aaa48a", + "compound": "fipronil", + "n": 50, + "wall_seconds": 2115.5855822919984, + "profile": { + "compound": "fipronil", + "conc_M": 1e-06, + "assay": "subgraph", + "graph": "/private/tmp/flylab-ladder/malecns_scale_50k.json", + "readout": "mean_hz", + "n": 50, + "seed": 0, + "alpha": 0.05, + "effect_floor": 0.0014003326665626462, + "effect_floor_absolute": 1e-06, + "effect_floor_frac": 0.01, + "effect_floor_relative": 0.0014003326665626462, + "effect_floor_binding": "relative", + "real_effect": 0.033669992129486404, + "real_vehicle": 0.14003326665626462, + "delta": 0.007001663332813231, + "delta_frac": 0.05, + "delta_scale": "0.05 x |vehicle readout| = 0.00700166", + "confirmatory": false, + "multiplicity_correction": null, + "design": "exploratory", + "p": { + "p_sign": 0.0196078431372549, + "p_weight": 0.0196078431372549, + "p_degree": 0.0196078431372549, + "p_ER": 0.0196078431372549, + "p_sign_wm": 0.0392156862745098 + }, + "q": { + "q_sign": null, + "q_weight": null, + "q_degree": null, + "q_ER": null, + "q_sign_wm": null + }, + "p_resolution": 0.0196078431372549, + "resolution_coarser_than_alpha": false, + "D": { + "z_sign": -2.548013646999591, + "z_weight": -4.32614963620229, + "z_degree": 15.680607013529865, + "z_ER": -15.52511665963148, + "z_sign_wm": -1.8735715007239808 + }, + "D_vector": [ + -2.548013646999591, + -4.32614963620229, + 15.680607013529865, + -15.52511665963148 + ], + "D_keys": [ + "z_sign", + "z_weight", + "z_degree", + "z_ER" + ], + "modes_run": [ + "sign_permute", + "weight_permute", + "rewire_degree_preserving", + "erdos_renyi", + "sign_permute_weight_matched" + ], + "modes": [ + { + "mode": "sign_permute", + "information_kept": "the real edges, the real weights and the unweighted NT histogram", + "rank": 3, + "real_effect": 0.033669992129486404, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": -2.548013646999591, + "null_mean": 0.0662972447492532, + "null_median": 0.06471705334271999, + "null_sd": 0.012804975616275433, + "n": 50, + "n_ok": 50, + "runtime_s": 33.87826041702647, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.031047061213233584, + "delta": 0.007001663332813231, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "weight_permute", + "information_kept": "the real edge list and the transmitters", + "rank": 2, + "real_effect": 0.033669992129486404, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": -4.32614963620229, + "null_mean": 0.0517570001909704, + "null_median": 0.05117535002692929, + "null_sd": 0.0041808558608623795, + "n": 50, + "n_ok": 50, + "runtime_s": 36.76907416700851, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.017505357897442883, + "delta": 0.007001663332813231, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "rewire_degree_preserving", + "information_kept": "every node's in/out degree and its transmitter", + "rank": 1, + "real_effect": 0.033669992129486404, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": 15.680607013529865, + "null_mean": 0.02149159519999089, + "null_median": 0.02152085605074027, + "null_sd": 0.0007766534113754332, + "n": 50, + "n_ok": 50, + "runtime_s": 1913.5263985830243, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.012149136078746132, + "delta": 0.007001663332813231, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "erdos_renyi", + "information_kept": "N, E, the weight histogram and the transmitter census", + "rank": 0, + "real_effect": 0.033669992129486404, + "p_two_sided": 0.0196078431372549, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": 25, + "resolution_limited": true, + "verdict_stable": false, + "z": -15.52511665963148, + "null_mean": 0.0695844017827434, + "null_median": 0.06993130625005678, + "null_sd": 0.0023133101309726002, + "n": 50, + "n_ok": 50, + "runtime_s": 65.27275216695853, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.036261314120570376, + "delta": 0.007001663332813231, + "p_used": 0.0196078431372549, + "alpha": 0.05, + "p_basis": "raw" + }, + { + "mode": "sign_permute_weight_matched", + "information_kept": "the real edges, the real weights, the NT histogram AND the share of total outgoing synaptic weight carried by each transmitter", + "rank": 3, + "real_effect": 0.033669992129486404, + "p_two_sided": 0.0392156862745098, + "p_resolution": 0.0196078431372549, + "beats_null": true, + "stabilised": false, + "n_stabilised": null, + "resolution_limited": false, + "verdict_stable": false, + "z": -1.8735715007239808, + "null_mean": 0.05343257107876521, + "null_median": 0.051876452918559204, + "null_sd": 0.010548078331487321, + "n": 50, + "n_ok": 50, + "verdict": "distinguishable", + "verdict_reason": "the empirical two-sided permutation p = 0.03922 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "distinguishable": true, + "within_tolerance": false, + "abs_gap_from_null_median": 0.0182064607890728, + "delta": 0.007001663332813231, + "p_used": 0.0392156862745098, + "alpha": 0.05, + "p_basis": "raw" + } + ], + "verdicts": { + "sign_permute": "distinguishable", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "distinguishable" + }, + "classification": { + "class": "topology-dependent", + "reason": "the effect is distinguishable from weight_permute and rewire_degree_preserving, so it needs the wiring pattern and not only the graph's size, degree sequence and transmitter census.", + "alpha": 0.05, + "effect_floor": 0.0014003326665626462, + "delta": 0.007001663332813231, + "mode_verdicts": { + "sign_permute": "distinguishable", + "weight_permute": "distinguishable", + "rewire_degree_preserving": "distinguishable", + "erdos_renyi": "distinguishable", + "sign_permute_weight_matched": "distinguishable" + }, + "mode_verdict_reasons": { + "sign_permute": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "weight_permute": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "rewire_degree_preserving": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "erdos_renyi": "the empirical two-sided permutation p = 0.01961 is at or below alpha = 0.05: the real graph and this graph model give different drug effects.", + "sign_permute_weight_matched": "the empirical two-sided permutation p = 0.03922 is at or below alpha = 0.05: the real graph and this graph model give different drug effects." + }, + "within_tolerance": { + "sign_permute": false, + "weight_permute": false, + "rewire_degree_preserving": false, + "erdos_renyi": false, + "sign_permute_weight_matched": false + }, + "distinguishable": { + "sign_permute": true, + "weight_permute": true, + "rewire_degree_preserving": true, + "erdos_renyi": true, + "sign_permute_weight_matched": true + }, + "structural_distinguishable": [ + "weight_permute", + "rewire_degree_preserving" + ], + "equivalent_modes": [], + "indeterminate_modes": [], + "n_indeterminate": 0, + "er_distinguishable": true, + "equivalence": "no mode was shown equivalent within tolerance", + "network_insensitive": false + }, + "class": "topology-dependent", + "stabilised": false, + "verdict_stable": false, + "n_stabilised": 25, + "runtime_s": 2115.5854126670165, + "label": "model_derived", + "warnings": [ + "Null models destroy the MaleCNS wiring on purpose: they are controls, not biology. Nothing here is a measurement from a living fly.", + "The headline statistic is the empirical two-sided permutation p, (k+1)/(n+1); with n shuffles its resolution (and floor) is 1/(n+1), so a p at the floor means 'not resolvable with this n', not 'p = 0'.", + "z is reported second and is not a probability: the null distribution on a degraded graph is usually not normal, and a large |z| can simply mean the shuffles collapsed onto one value.", + "model_derived: these numbers describe this simulation and its teaching EC50 library, not a measured drug effect.", + "The dependence class is a documented label built on the per-mode verdicts; within one cell no correction is applied across the four null modes, and the alpha threshold is a reporting convention.", + "'Not distinguishable' is NOT 'equivalent': only a mode whose gap from the null median is below the prespecified margin delta is reported as equivalent_within_tolerance. Everything else that fails to reject is indeterminate and is bounded by the permutation resolution 1/(n+1).", + "sign_permute keeps the transmitter histogram exactly, so a small |z| means the drug effect follows global E/I balance rather than the identity of the cholinergic cells.", + "shuffles were run on 4 processes; shuffle i is seeded by (seed, i) alone, so the draws are identical to a serial run.", + "p is at its resolution floor 0.0196: with 50 shuffles this is the smallest p obtainable, so report it as 'p <= floor', not as an exact value.", + "the permutation p had not stabilised at this n (it moved by 0.057 between checkpoints, tolerance 0.020); the verdict at alpha also changed between checkpoints. Increase n before quoting the value.", + "weight_permute keeps topology and transmitters, so it isolates the pairing between synapse counts and wiring.", + "rewire_degree_preserving keeps every node's in/out degree and transmitter, so it isolates the wiring pattern itself.", + "erdos_renyi keeps only N, E and the weight histogram; it is the weakest null and should be the easiest to beat.", + "exploratory: this profile is not one of the prespecified confirmatory tests (imidacloprid, fipronil at n >= 1000) and carries no multiplicity correction of its own. Counting verdicts over many such profiles needs the component-test-adjusted dependence_landscape instead." + ], + "necessary_information_level": { + "level": "real_connectome", + "mode": "real_connectome", + "rank": 4, + "keeps": "the MaleCNS cut itself", + "destroys": "nothing", + "p_two_sided": null, + "verdict": "distinguishable", + "equivalent_within_tolerance": false, + "delta": 0.007001663332813231, + "description": "every degradation of the graph was distinguishable from the real one, so this prediction needs the real MaleCNS wiring.", + "class": "topology-dependent", + "warnings": [] + } + } + } + ], + "status": "complete", + "completed_at_utc": "2026-09-24T17:49:50.849522+00:00", + "analysis_limitations": [ + "Exploratory: all profiles use fewer than the prespecified n=1000 confirmatory threshold.", + "No multiplicity adjustment was made across the eight cut-by-compound profiles; interpret the profile family descriptively.", + "The historical 5k-50k cut hashes and source-file hashes were not saved. The rebuilt 1k node and edge lists match the committed cut exactly, and larger node/edge counts match the saved T27 fingerprints, but byte identity to the historical larger cuts is unverified.", + "The saved papers/scale_study.json and T27 manuscript results were not changed." + ], + "reconstruction_validation": { + "scale_1k_nodes_equal_committed": true, + "scale_1k_edges_equal_committed": true, + "larger_node_edge_counts_match_saved_T27": true, + "historical_larger_cut_byte_identity_verified": false + }, + "replicate_family": { + "profiles": 8, + "permutation_adjustment_across_profiles": "none", + "status": "exploratory" + } +} diff --git a/data/literature/README.md b/data/literature/README.md index c4e0524..13c962e 100644 --- a/data/literature/README.md +++ b/data/literature/README.md @@ -82,9 +82,32 @@ expressed in the midbrain but α1/α5/α6/α7 in far more cells than α2/α3/α4 β1 in more than twice as many cells as β2 (Croset 2018) — meaning `insect_nAChR` is not one receptor and subunit-selective compounds reach only partly overlapping cell sets; and GluClα is broadly expressed in the VNC. -**Motor neurons are a confirmed gap**: no cited source reports Rdl or nAChR -expression for identified adult leg or labellar motor neurons, which is -recorded in a `motor_neuron_gap` block with a suggested next step. +**Motor-neuron receptor fractions have not been added to this dataset.** We +reprocessed the four adult VNC count matrices deposited in GEO GSE141807 into +VNC-wide gene-detection fractions after the paper's QC filters; see +[`vnc_receptor_detection_GSE141807.yaml`](vnc_receptor_detection_GSE141807.yaml). +These summaries have no cell-class assignments, so they do not estimate +motor-neuron fractions or receptor-complex prevalence. One replicate retains +two more cells than the paper's reported count under its published thresholds; +the discrepancy is recorded rather than adjusted. To reproduce the summary, +download `GSE141807_RAW.tar` from GEO and run +`python scripts/reanalyze_vnc_expression.py GSE141807_RAW.tar`; the script checks +the archive SHA-256 before processing. The matrices do not identify +MN9: McKellar et al. place the identified proboscis motor-neuron arbors in the +brain subesophageal zone. The `motor_neuron_gap` block records this scope and a +next step. Keep uniform gains until a defensible cell-class match is available. + +The GEO samples expose replicate-specific barcode matrices, but no published +barcode-to-cell-class table. Allen et al. provide annotations for 120 clusters +in [Figure 1 source data 1](https://cdn.elifesciences.org/articles/54074/elife-54074-fig1-data1-v1.xlsx) +and [source data 2](https://cdn.elifesciences.org/articles/54074/elife-54074-fig1-data2-v1.xlsx); +these files do not map individual barcodes to clusters. Recovering barcode-level +labels requires rerunning the authors' [Seurat analysis](https://github.com/aaron-allen/VNC_scRNAseq) +or obtaining its cell-level object, then joining cluster IDs to the paper's +cluster summaries. That still would not make predicted hemilineage a motor-neuron +label. Any reconstructed key must include both the GEO sample accession and the +barcode, since barcodes alone are not a safe join across replicates. These adult +VNC data also do not cover MN9's brain/SEZ arbors. **`mixtures.yaml` — 4 reference models, 7 empirical entries.** Bliss independence and Loewe additivity are written out as equations with variable @@ -281,7 +304,7 @@ Every source used across the six files, once each. - Chen W, Gu X, Yang YT, Batterham P, Perry T (2022) Dual nicotinic acetylcholine receptor subunit gene knockouts reveal limits to functional redundancy. *Pestic Biochem Physiol* 184:105118. DOI 10.1016/j.pestbp.2022.105118. PMID 35715057. - French AS, Sellier MJ, Ali Agha M, Guigue A, Chabaud MA, Reeb PD, Mitra A, Grau Y, Soustelle L, Marion-Poll F (2015) Dual mechanism for bitter avoidance in *Drosophila*. *J Neurosci* 35(9):3990–4004. DOI 10.1523/JNEUROSCI.1312-14.2015. PMID 25740527. - Gargano JW, Martin I, Bhandari P, Grotewiel MS (2005) Rapid iterative negative geotaxis (RING): a new method for assessing age-related locomotor decline in *Drosophila*. *Exp Gerontol* 40(5):386–395. DOI 10.1016/j.exger.2005.02.005. PMID 15919590. -- Lee HY, Zhou L, Ghanta S, Asplund M, Zhu C, et al. (2014) Mechanisms of naturally evolved ethanol resistance in *Drosophila melanogaster*. *J Exp Biol*. DOI 10.1242/jeb.110510. *(cited only as the source of a null in `fly_pharmacokinetics.yaml`)* +- Fry JD (2014) Mechanisms of naturally evolved ethanol resistance in *Drosophila melanogaster*. *J Exp Biol* 217(22):3996–4003. DOI 10.1242/jeb.110510. *(cited only as the source of a null in `fly_pharmacokinetics.yaml`)* - Martelli F, Zhongyuan Z, Wang J, Wong CO, Karagas NE, Roessner U, Rupasinghe T, Venkatachalam K, Perry T, Bellen HJ, Batterham P (2020) Low doses of the neonicotinoid insecticide imidacloprid induce ROS triggering neurological and metabolic impairments in *Drosophila*. *Proc Natl Acad Sci USA* 117(41):25840–25850. DOI 10.1073/pnas.2011828117. PMID 32989137. ### Cited in `flylab/pharm/library.yaml` and re-checked here diff --git a/data/literature/fly_pharmacokinetics.yaml b/data/literature/fly_pharmacokinetics.yaml index dd1d41b..bfe77d6 100644 --- a/data/literature/fly_pharmacokinetics.yaml +++ b/data/literature/fly_pharmacokinetics.yaml @@ -196,7 +196,7 @@ elimination_and_metabolism: notes: >- NOT SOURCED. Despite a large Drosophila ethanol literature, no publication giving an elimination rate constant or half-life for internal - ethanol in adult flies was verified. Lee HY et al. (2014) J Exp Biol + ethanol in adult flies was verified. Fry JD (2014) J Exp Biol (DOI 10.1242/jeb.110510) measure internal ethanol concentrations and show that tropical African flies accumulate 2-3x more internal ethanol than European flies under the same vapour exposure - a difference in net diff --git a/data/literature/receptor_expression_by_class.yaml b/data/literature/receptor_expression_by_class.yaml index bcc1d28..23b07e9 100644 --- a/data/literature/receptor_expression_by_class.yaml +++ b/data/literature/receptor_expression_by_class.yaml @@ -40,7 +40,10 @@ sources: allen_2020: citation: "Allen AM, Neville MC, Birtles S, et al. (2020) A single-cell transcriptomic atlas of the adult Drosophila ventral nerve cord. eLife 9:e54074. DOI 10.7554/eLife.54074. PMID 32314735. PMCID PMC7173974" scale: "26,768 high-quality cells; 120 clusters (110 neuronal); >100 transcriptionally distinct cell types" - covers: "adult VNC" + covers: "adult VNC; public primary count data in GEO GSE141807" + mckellar_2020: + citation: "McKellar CE, Siwanowicz I, Dickson BJ, Simpson JH (2020) Controlling motor neurons of every muscle for fly proboscis reaching. eLife 9:e54978. DOI 10.7554/eLife.54978. PMID 32584254. PMCID PMC7316511" + covers: "anatomy of identified adult proboscis motor neurons, including their arbors in the brain's subesophageal zone" croset_2018: citation: "Croset V, Treiber CD, Waddell S (2018) Cellular diversity in the Drosophila midbrain revealed by single-cell transcriptomics. eLife 7:e34550. DOI 10.7554/eLife.34550. PMID 29671739. PMCID PMC5927767" scale: "19,260 cells sequenced, 10,286 high-quality (>=800 UMI); authors estimate ~45,000 cells in the fly midbrain" @@ -348,12 +351,29 @@ by_transmitter_class: motor_neuron_gap: status: not_sourced description: >- - The task asked specifically for Rdl and nAChR expression in leg and labellar - motor neurons. No source among Fly Cell Atlas, Davie 2018, Allen 2020, - Croset 2018 or Konstantinides 2018 was found that reports receptor - expression for an identified motor neuron population in the adult. Allen et - al. 2020 is the closest (adult VNC, contains leg motor neurons) but the - retrieved text names no motor-neuron cluster with receptor expression. + No expression fractions for identified adult motor-neuron classes have been + added to this dataset. Allen et al. 2020 deposited primary adult VNC + single-cell count data (GEO GSE141807); those matrices have now been + reprocessed into descriptive, VNC-wide gene-detection fractions after the + published QC filters (see `vnc_receptor_detection_GSE141807.yaml`). The + available summary has no cell-class assignments and cannot estimate + motor-neuron-specific fractions. The VNC dataset does not itself identify MN9: + McKellar et al. 2020 place the proboscis motor-neuron arbors in the brain's + subesophageal zone. MN9 therefore still needs a brain/SEZ data source with + a defensible cell-type match or a targeted expression assay. + public_data_lead: + source_key: allen_2020 + accession: GSE141807 + status: reanalyzed_vnc_wide_cell_classes_unresolved + scope: adult VNC, not the brain subesophageal zone containing MN9 + summary_file: vnc_receptor_detection_GSE141807.yaml + processing_note: >- + Per-gene and target-family detection fractions were recomputed from all + four deposited replicates after the published QC filters, with sample and + sex composition, one-UMI detection, dropout caveat and a small unresolved + two-cell replicate discrepancy recorded. The pooled VNC-wide results are + not cell-class or receptor-complex prevalence estimates. An atlas zero is + not evidence of receptor absence in a cell. consequence_for_flylab: >- FlyLab currently patches g_ACh and g_GABA at MN9 as part of a whole-graph gain. There is no published per-cell receptor evidence justifying a @@ -361,10 +381,12 @@ motor_neuron_gap: stay, and any notebook that patches motor neurons should keep a warning saying the receptor complement of those cells is unmeasured. suggested_next_step: >- - Query the Fly Cell Atlas VNC/head snRNA-seq directly (SCope or cellxgene) - for nAChRalpha1-7, nAChRbeta1-3, Rdl, GluClalpha, Ace and Oct receptor genes - restricted to annotated motor neuron cell types, and record the resulting - fractions here as literature_value with the atlas version pinned. + Determine whether adult VNC motor-neuron labels can be defensibly joined to + the reprocessed GEO barcodes, then report class-specific subunit detection + fractions with the atlas version, sample composition, detection rule and + dropout treatment pinned. Resolve MN9 separately with a brain/SEZ atlas + that supports an identified cell-type match or a targeted expression assay; + do not transfer a generic VNC motor-neuron value to MN9. not_sourced: - "Per-subunit fractions of cells expressing each nAChR alpha/beta subunit (only ordinal statements retrieved)." diff --git a/data/literature/vnc_receptor_detection_GSE141807.yaml b/data/literature/vnc_receptor_detection_GSE141807.yaml new file mode 100644 index 0000000..d5d57c6 --- /dev/null +++ b/data/literature/vnc_receptor_detection_GSE141807.yaml @@ -0,0 +1,489 @@ +schema_version: 1 +source: + accession: GSE141807 + archive: GSE141807_RAW.tar + archive_sha256: bda31982a53a432bb4712ebdaf11ac7d8b9e7be580f72c2acaf7a58912e64daf + geo_url: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE141807 + paper_url: https://elifesciences.org/articles/54074 +method: + scope: adult VNC-wide transcript detection; cells are not assigned to cell classes + qc_thresholds_from_paper: + nGene_gt: 200 + nUMI_gt: 1200 + nUMI_lte: 10000 + prop_mito_lte: 0.15 + detection_rule: one or more UMI for the gene in a QC-passing barcode + dropout_note: a zero count is not evidence that a cell does not express the gene + not_used_for: MaleCNS superclass gains, receptor-complex inference, or MN9 expression +cells_after_qc_total: 26770 +published_cells_after_qc_total: 26768 +all_replicate_cell_counts_match_paper: false +cell_count_reproduction_note: Three replicate counts match the paper. GSM4213598 retains + 6,524 cells versus 6,522 reported (difference +2) under the cited QC thresholds; + this small discrepancy is unresolved and was not adjusted away. +replicates: + GSM4213596: + sample_id: GSM4213596 + sex: female + cells_in_matrix: 30000 + genes_in_matrix: 12253 + cells_after_published_qc: 2590 + published_cells_after_qc: 2590 + cell_count_difference: 0 + mitochondrial_genes: + - mt:ATPase6 + - mt:ATPase8 + - mt:CoI + - mt:CoII + - mt:CoIII + - mt:Cyt-b + - mt:ND1 + - mt:ND2 + - mt:ND3 + - mt:ND4 + - mt:ND4L + - mt:ND5 + - mt:ND6 + - mt:lrRNA + - mt:srRNA + - mt:tRNA:Arg-TCG + - mt:tRNA:Cys-GCA + - mt:tRNA:Gly-TCC + - mt:tRNA:Leu-TAA + - mt:tRNA:Lys-CTT + - mt:tRNA:Ser-GCT + - mt:tRNA:Ser-TGA + - mt:tRNA:Val-TAC + targets: + insect_nAChR: + n_genes_matched: 10 + genes: + nAChRalpha1: + n_detected: 1206 + fraction_detected: 0.46563706563706564 + nAChRalpha2: + n_detected: 167 + fraction_detected: 0.06447876447876448 + nAChRalpha3: + n_detected: 1540 + fraction_detected: 0.5945945945945946 + nAChRalpha4: + n_detected: 562 + fraction_detected: 0.21698841698841698 + nAChRalpha5: + n_detected: 1858 + fraction_detected: 0.7173745173745174 + nAChRalpha6: + n_detected: 2154 + fraction_detected: 0.8316602316602316 + nAChRalpha7: + n_detected: 1861 + fraction_detected: 0.7185328185328186 + nAChRbeta1: + n_detected: 2348 + fraction_detected: 0.9065637065637066 + nAChRbeta2: + n_detected: 1302 + fraction_detected: 0.5027027027027027 + nAChRbeta3: + n_detected: 1 + fraction_detected: 0.0003861003861003861 + n_cells_with_any_matched_gene_detected: 2568 + fraction_cells_with_any_matched_gene_detected: 0.9915057915057915 + insect_RDL: + n_genes_matched: 1 + genes: + Rdl: + n_detected: 2554 + fraction_detected: 0.9861003861003861 + n_cells_with_any_matched_gene_detected: 2554 + fraction_cells_with_any_matched_gene_detected: 0.9861003861003861 + insect_GluCl: + n_genes_matched: 1 + genes: + GluClalpha: + n_detected: 1223 + fraction_detected: 0.4722007722007722 + n_cells_with_any_matched_gene_detected: 1223 + fraction_cells_with_any_matched_gene_detected: 0.4722007722007722 + insect_OctR: + n_genes_matched: 4 + genes: + Oct-TyrR: + n_detected: 195 + fraction_detected: 0.07528957528957529 + Octbeta1R: + n_detected: 147 + fraction_detected: 0.05675675675675676 + Octbeta2R: + n_detected: 796 + fraction_detected: 0.3073359073359073 + Octbeta3R: + n_detected: 1094 + fraction_detected: 0.4223938223938224 + n_cells_with_any_matched_gene_detected: 1614 + fraction_cells_with_any_matched_gene_detected: 0.6231660231660232 + insect_AChE: + n_genes_matched: 2 + genes: + Ace: + n_detected: 1270 + fraction_detected: 0.49034749034749037 + Acer: + n_detected: 184 + fraction_detected: 0.07104247104247104 + n_cells_with_any_matched_gene_detected: 1364 + fraction_cells_with_any_matched_gene_detected: 0.5266409266409267 + GSM4213597: + sample_id: GSM4213597 + sex: female + cells_in_matrix: 30000 + genes_in_matrix: 13525 + cells_after_published_qc: 9060 + published_cells_after_qc: 9060 + cell_count_difference: 0 + mitochondrial_genes: + - mt:ATPase6 + - mt:ATPase8 + - mt:CoI + - mt:CoII + - mt:CoIII + - mt:Cyt-b + - mt:ND1 + - mt:ND2 + - mt:ND3 + - mt:ND4 + - mt:ND4L + - mt:ND5 + - mt:ND6 + - mt:lrRNA + - mt:srRNA + - mt:tRNA:Ala-TGC + - mt:tRNA:Arg-TCG + - mt:tRNA:Cys-GCA + - mt:tRNA:Gly-TCC + - mt:tRNA:His-GTG + - mt:tRNA:Leu-TAA + - mt:tRNA:Leu-TAG + - mt:tRNA:Lys-CTT + - mt:tRNA:Met-CAT + - mt:tRNA:Phe-GAA + - mt:tRNA:Ser-GCT + - mt:tRNA:Ser-TGA + - mt:tRNA:Trp-TCA + - mt:tRNA:Tyr-GTA + - mt:tRNA:Val-TAC + targets: + insect_nAChR: + n_genes_matched: 10 + genes: + nAChRalpha1: + n_detected: 3415 + fraction_detected: 0.3769315673289183 + nAChRalpha2: + n_detected: 2678 + fraction_detected: 0.29558498896247243 + nAChRalpha3: + n_detected: 4842 + fraction_detected: 0.5344370860927152 + nAChRalpha4: + n_detected: 2312 + fraction_detected: 0.25518763796909494 + nAChRalpha5: + n_detected: 6234 + fraction_detected: 0.6880794701986755 + nAChRalpha6: + n_detected: 7333 + fraction_detected: 0.8093818984547462 + nAChRalpha7: + n_detected: 6223 + fraction_detected: 0.6868653421633554 + nAChRbeta1: + n_detected: 8021 + fraction_detected: 0.8853200883002208 + nAChRbeta2: + n_detected: 4050 + fraction_detected: 0.4470198675496689 + nAChRbeta3: + n_detected: 18 + fraction_detected: 0.001986754966887417 + n_cells_with_any_matched_gene_detected: 8867 + fraction_cells_with_any_matched_gene_detected: 0.9786975717439294 + insect_RDL: + n_genes_matched: 1 + genes: + Rdl: + n_detected: 8741 + fraction_detected: 0.9647902869757174 + n_cells_with_any_matched_gene_detected: 8741 + fraction_cells_with_any_matched_gene_detected: 0.9647902869757174 + insect_GluCl: + n_genes_matched: 1 + genes: + GluClalpha: + n_detected: 4989 + fraction_detected: 0.5506622516556291 + n_cells_with_any_matched_gene_detected: 4989 + fraction_cells_with_any_matched_gene_detected: 0.5506622516556291 + insect_OctR: + n_genes_matched: 4 + genes: + Oct-TyrR: + n_detected: 993 + fraction_detected: 0.10960264900662252 + Octbeta1R: + n_detected: 901 + fraction_detected: 0.09944812362030905 + Octbeta2R: + n_detected: 2614 + fraction_detected: 0.28852097130242826 + Octbeta3R: + n_detected: 2904 + fraction_detected: 0.3205298013245033 + n_cells_with_any_matched_gene_detected: 5276 + fraction_cells_with_any_matched_gene_detected: 0.5823399558498896 + insect_AChE: + n_genes_matched: 2 + genes: + Ace: + n_detected: 4975 + fraction_detected: 0.5491169977924945 + Acer: + n_detected: 599 + fraction_detected: 0.06611479028697571 + n_cells_with_any_matched_gene_detected: 5207 + fraction_cells_with_any_matched_gene_detected: 0.5747240618101546 + GSM4213598: + sample_id: GSM4213598 + sex: male + cells_in_matrix: 30000 + genes_in_matrix: 13261 + cells_after_published_qc: 6524 + published_cells_after_qc: 6522 + cell_count_difference: 2 + mitochondrial_genes: + - mt:ATPase6 + - mt:ATPase8 + - mt:CoI + - mt:CoII + - mt:CoIII + - mt:Cyt-b + - mt:ND1 + - mt:ND2 + - mt:ND3 + - mt:ND4 + - mt:ND4L + - mt:ND5 + - mt:ND6 + - mt:lrRNA + - mt:srRNA + - mt:tRNA:Ala-TGC + - mt:tRNA:Arg-TCG + - mt:tRNA:Gly-TCC + - mt:tRNA:His-GTG + - mt:tRNA:Leu-TAA + - mt:tRNA:Lys-CTT + - mt:tRNA:Ser-GCT + - mt:tRNA:Ser-TGA + - mt:tRNA:Val-TAC + targets: + insect_nAChR: + n_genes_matched: 10 + genes: + nAChRalpha1: + n_detected: 2516 + fraction_detected: 0.38565297363580625 + nAChRalpha2: + n_detected: 811 + fraction_detected: 0.12431023911710606 + nAChRalpha3: + n_detected: 3456 + fraction_detected: 0.5297363580625383 + nAChRalpha4: + n_detected: 1395 + fraction_detected: 0.21382587369711833 + nAChRalpha5: + n_detected: 4391 + fraction_detected: 0.6730533415082771 + nAChRalpha6: + n_detected: 5302 + fraction_detected: 0.8126916002452483 + nAChRalpha7: + n_detected: 4509 + fraction_detected: 0.6911404046597179 + nAChRbeta1: + n_detected: 5838 + fraction_detected: 0.8948497854077253 + nAChRbeta2: + n_detected: 2911 + fraction_detected: 0.4461986511342734 + nAChRbeta3: + n_detected: 3 + fraction_detected: 0.0004598405885959534 + n_cells_with_any_matched_gene_detected: 6451 + fraction_cells_with_any_matched_gene_detected: 0.9888105456774985 + insect_RDL: + n_genes_matched: 1 + genes: + Rdl: + n_detected: 6381 + fraction_detected: 0.9780809319435929 + n_cells_with_any_matched_gene_detected: 6381 + fraction_cells_with_any_matched_gene_detected: 0.9780809319435929 + insect_GluCl: + n_genes_matched: 1 + genes: + GluClalpha: + n_detected: 3022 + fraction_detected: 0.4632127529123237 + n_cells_with_any_matched_gene_detected: 3022 + fraction_cells_with_any_matched_gene_detected: 0.4632127529123237 + insect_OctR: + n_genes_matched: 4 + genes: + Oct-TyrR: + n_detected: 417 + fraction_detected: 0.06391784181483752 + Octbeta1R: + n_detected: 480 + fraction_detected: 0.07357449417535254 + Octbeta2R: + n_detected: 1824 + fraction_detected: 0.2795830778663397 + Octbeta3R: + n_detected: 2080 + fraction_detected: 0.31882280809319435 + n_cells_with_any_matched_gene_detected: 3591 + fraction_cells_with_any_matched_gene_detected: 0.5504291845493562 + insect_AChE: + n_genes_matched: 2 + genes: + Ace: + n_detected: 3105 + fraction_detected: 0.47593500919681175 + Acer: + n_detected: 400 + fraction_detected: 0.061312078479460456 + n_cells_with_any_matched_gene_detected: 3282 + fraction_cells_with_any_matched_gene_detected: 0.503065603923973 + GSM4213599: + sample_id: GSM4213599 + sex: male + cells_in_matrix: 30000 + genes_in_matrix: 14232 + cells_after_published_qc: 8596 + published_cells_after_qc: 8596 + cell_count_difference: 0 + mitochondrial_genes: + - mt:ATPase6 + - mt:ATPase8 + - mt:CoI + - mt:CoII + - mt:CoIII + - mt:Cyt-b + - mt:ND1 + - mt:ND2 + - mt:ND3 + - mt:ND4 + - mt:ND4L + - mt:ND5 + - mt:ND6 + - mt:lrRNA + - mt:srRNA + - mt:tRNA:Ala-TGC + - mt:tRNA:Arg-TCG + - mt:tRNA:Cys-GCA + - mt:tRNA:Gln-TTG + - mt:tRNA:Glu-TTC + - mt:tRNA:Gly-TCC + - mt:tRNA:His-GTG + - mt:tRNA:Leu-TAA + - mt:tRNA:Lys-CTT + - mt:tRNA:Phe-GAA + - mt:tRNA:Ser-GCT + - mt:tRNA:Ser-TGA + - mt:tRNA:Trp-TCA + - mt:tRNA:Tyr-GTA + - mt:tRNA:Val-TAC + targets: + insect_nAChR: + n_genes_matched: 10 + genes: + nAChRalpha1: + n_detected: 3213 + fraction_detected: 0.3737785016286645 + nAChRalpha2: + n_detected: 2575 + fraction_detected: 0.29955793392275476 + nAChRalpha3: + n_detected: 4273 + fraction_detected: 0.49709167054443926 + nAChRalpha4: + n_detected: 2117 + fraction_detected: 0.24627733829688228 + nAChRalpha5: + n_detected: 5985 + fraction_detected: 0.6962540716612378 + nAChRalpha6: + n_detected: 6972 + fraction_detected: 0.8110749185667753 + nAChRalpha7: + n_detected: 5948 + fraction_detected: 0.6919497440670079 + nAChRbeta1: + n_detected: 7610 + fraction_detected: 0.885295486272685 + nAChRbeta2: + n_detected: 3731 + fraction_detected: 0.43403908794788276 + nAChRbeta3: + n_detected: 15 + fraction_detected: 0.0017449976733364355 + n_cells_with_any_matched_gene_detected: 8455 + fraction_cells_with_any_matched_gene_detected: 0.9835970218706375 + insect_RDL: + n_genes_matched: 1 + genes: + Rdl: + n_detected: 8314 + fraction_detected: 0.967194043741275 + n_cells_with_any_matched_gene_detected: 8314 + fraction_cells_with_any_matched_gene_detected: 0.967194043741275 + insect_GluCl: + n_genes_matched: 1 + genes: + GluClalpha: + n_detected: 4664 + fraction_detected: 0.542577943229409 + n_cells_with_any_matched_gene_detected: 4664 + fraction_cells_with_any_matched_gene_detected: 0.542577943229409 + insect_OctR: + n_genes_matched: 4 + genes: + Oct-TyrR: + n_detected: 864 + fraction_detected: 0.10051186598417869 + Octbeta1R: + n_detected: 858 + fraction_detected: 0.09981386691484412 + Octbeta2R: + n_detected: 2478 + fraction_detected: 0.28827361563517917 + Octbeta3R: + n_detected: 2637 + fraction_detected: 0.3067705909725454 + n_cells_with_any_matched_gene_detected: 4907 + fraction_cells_with_any_matched_gene_detected: 0.5708469055374593 + insect_AChE: + n_genes_matched: 2 + genes: + Ace: + n_detected: 4936 + fraction_detected: 0.5742205677059097 + Acer: + n_detected: 568 + fraction_detected: 0.0660772452303397 + n_cells_with_any_matched_gene_detected: 5159 + fraction_cells_with_any_matched_gene_detected: 0.6001628664495114 +interpretation: Descriptive VNC-wide RNA detection only. Detection fractions are not + receptor complex prevalence or cell-type-specific expression fractions. diff --git a/docs/INTERPRETATION.md b/docs/INTERPRETATION.md index b1306f3..d2be818 100644 --- a/docs/INTERPRETATION.md +++ b/docs/INTERPRETATION.md @@ -265,38 +265,60 @@ On `taste_motor`, `rewire_degree_preserving` reaches the resolution floor (p = 0.0050 at n = 200) and the cell is topology-dependent. On `named`, no structure-preserving null can be told apart at all. -### And again with size - -The repository now carries a scale ladder — the same connectome, cut at 1000, -5000, 10 000, 25 000 and 50 000 cells (`flylab/maps/ladder.py`, -`flylab.analysis.scale.available_cuts()`). Walking a dependence profile up that -ladder moves the verdict again: - -> At about 1100 cells imidacloprid is composition-dominated. At about 5000 cells -> the same compound, at the same concentration, on the same engine, is -> distinguishable from **every** null — including the weight-matched transmitter -> null, the hardest rung on the ladder. - -At the time of writing, that is an in-flight measurement, not a committed -number. `papers/results.json` records `scale_study_status: "not yet run"`, and -its `scale_study_statement` says so in words: - -> Where the verdict settles with the size of the cut has not been measured at -> this commit: the scale ladder's rungs on disk are scale_1k, scale_5k, -> scale_10k, scale_25k, scale_50k and the study over the full ladder had not -> been run. - -**Where the verdict settles — whether it settles at all before the whole CNS — -is an open question currently being measured.** When the study lands, the -per-rung verdicts come from `flylab.analysis.scale.dependence_vs_scale()` and -`verdict_stability()`, and the shipped numbers from -`papers/results.json` (`scale_study_settled`, -`scale_study_settled_clean_ladder`, `scale_study_statement`). Quote those, not -this paragraph. - -A related limit is already committed: `papers/tables/T26_scale_frontier.md` puts -the largest feasible dependence profile at `scale_50k`, and a profile on the -whole 165 122-cell CNS at about 59 hours on one core — outside the feasible +### The saved scale ladder + +T27 records a scale study across seven cuts; its values are imported from the +saved `papers/scale_study.json`. The versioned checkout contains the 1k rung +but not the larger 5k–50k cuts or the source Feather files. On 2026-09-24, the +larger cuts were rebuilt from public MaleCNS v1.0 inputs for a separate +exploratory follow-up; see +[`data/derived/scale_study_high_power_2026-09-24.json`](../data/derived/scale_study_high_power_2026-09-24.json). +The rebuilt 1k nodes and edges match the committed cut exactly, and the larger +rungs match T27's node/edge counts. Historical input and cut hashes were not +recorded, so byte identity to the original 5k–50k cuts cannot be verified. + +On the saved `scale_1k` cut (1000 nodes, mean degree 22.86), imidacloprid is +topology-dependent even though it has fewer nodes than `named` (1126 nodes, +mean degree 1.21), which is composition-dominated. This comparison is +consistent with recurrence tracking the difference, but does not isolate a +causal structural feature: construction rule and graph structure differ too. + +At `scale_5k` and `scale_10k`, the imidacloprid effect is distinguishable from +each of the three ranked edge/wiring nulls (`weight_permute`, +`rewire_degree_preserving`, `erdos_renyi`). At 5k both transmitter-label checks +are also distinguishable; at 10k the plain `sign_permute` check is +indeterminate. At 25k and 50k the three ranked edge/wiring nulls remain +distinguishable, while the transmitter-label checks are not both +distinguishable. Those top rungs use n = 20 (p resolution 0.0476), so each +rejection is the smallest this test can express and confirms rather than +establishes the pattern. The orthogonal transmitter-label checks are separate +from the ranked necessary-information ladder. + +T27 remains the original saved result; the 2026-09-24 follow-up is a separate +run and does not replace it. The follow-up used n = 100 at 5k and n = 50 at +10k–50k for imidacloprid and fipronil. All eight profiles are classified +topology-dependent. For each profile, the three ranked structural nulls +(`weight_permute`, `rewire_degree_preserving`, `erdos_renyi`) are +distinguishable at the smallest attainable p: 0.0099 at 5k and 0.0196 at +10k–50k. The transmitter-label checks vary separately: at 50k, imidacloprid's +plain label permutation is equivalent within the margin (p = 0.843), while its +weight-matched check is indeterminate (p = 0.529); both checks are +distinguishable for fipronil (p = 0.0196 and 0.0392). These eight profiles are +exploratory because n is below the confirmatory threshold of 1000, and no +multiplicity adjustment was made across the cut-by-compound family. Do not +promote them to confirmatory evidence or fold them into T27 without a reviewed +paper rerender. + +The degree-preserving edge-swap sampler remains the cost bottleneck. Observed +local wall time per profile ranged from 7.4–7.6 minutes at 5k, 8.4–8.7 minutes +at 10k, 24.9–27.4 minutes at 25k, and 35.3–41.4 minutes at 50k, using four +workers. Any faster sampler must retain a simple degree-preserving graph and a +valid, seeded permutation procedure; do not extrapolate a new verdict from +partial rewires or a changed null definition. + +A related limit is already committed: `papers/tables/T26_scale_frontier.md` +puts the largest feasible dependence profile at `scale_50k`, and a profile on +the whole 165 122-cell CNS at about 59 hours on one core — outside the feasible envelope. So "run it on the whole connectome and settle it" is not currently an option, which is exactly why the caveat has to travel with the verdict. @@ -339,9 +361,10 @@ Repeat any negative topology result on a denser cut before generalising it. ``` **What it does not mean.** Neither verdict is "the truth about imidacloprid". -Both are true statements about a readout on a substrate, and the interesting -question is not which one is right — it is *where the verdict settles as the cut -grows*, which is the thing the scale study is measuring. +Both are statements about a readout on a particular substrate. T27 records how +the verdict changes across its saved ladder; absent larger cut files prevent an +independent rerun here, and the cross-cut comparison does not prove recurrence +is the cause. **The misreading to avoid.** Referee item **B3**: the paper generalised from the `named` cut anyway. Never write "imidacloprid's effect does not need the @@ -351,14 +374,15 @@ distinguishable from the degree-preserving null ensemble at n = 1000" — and if you are making a negative claim, repeat it on a denser cut before you believe it. -NOVELTY.md now treats this reversal as the result rather than a caveat ("The -verdict is substrate-dependent, and that is now the result"), and withdraws -"most predictions do not need the connectome" along with it. The instrument is -not what is in doubt: `synthetic_cut`, `ladder_recovery` and `ladder_power` in -`dependence.py` plant a recurrent loop of known strength in a synthetic graph of -matched size, density and composition, check that the ladder recovers it and -that the unplanted control is not called, and map detection rate against effect -size and permutation count. The reversal is about the graphs. +NOVELTY.md now treats the two-cut difference as a substrate-conditional result +and withdraws "most predictions do not need the connectome". The instrument +also has a limited planted-cycle check: `synthetic_cut`, `ladder_recovery` and +`ladder_power` test recovery for one planted mechanism and report a control +count, but do not establish general power for biological endpoints. The +`named` versus `taste_motor` comparison uses the same n = 1000 effort; the T27 +scale rungs use different, often smaller permutation counts. So the two-cut +contrast is conditional on those extracts and settings, while the scale trend +is descriptive and cannot rule out power differences. **Always name four things with a dependence verdict** — the compound, the concentration, the readout and the cut — **and two numbers about the cut**: its @@ -774,8 +798,11 @@ particular: *predicted* consensus transmitters. A wrong prediction flips a synapse, and the drug patch follows the label rather than the biology. - **No per-cell receptor expression.** The gain is applied uniformly to every - cell of a transmitter class; about 20% of cells are mapped. A cell that does - not express the receptor is patched exactly like one that does. + cell of a transmitter class. About 20.5% of graph-node × receptor-key pairs + across the two committed cuts have a coarse class-level annotation; this is + annotation availability, not the share of cells measured to express a + receptor. A cell that does not express the receptor is patched exactly like + one that does. And there is deliberately **no single confidence score** anywhere in the bench. The layers fail independently, and one blended number would hide which of them diff --git a/docs/NOVELTY.md b/docs/NOVELTY.md index 17c8ea5..e886c87 100644 --- a/docs/NOVELTY.md +++ b/docs/NOVELTY.md @@ -75,17 +75,34 @@ over: Only the equivalent-within-tolerance verdict licenses the word "reproduces", and even then it means "within the prespecified margin", never "the shuffle gave the same effect". Check the verdict against a run rather than assuming: for imidacloprid at the headline dose and n = 1000, the weight, degree-preserving and weight-matched transmitter modes fall inside the margin and report **equivalent_within_tolerance**, while the plain transmitter permutation reports **indeterminate** (gap 2.65 Hz against a margin of 0.33 Hz) and the Erdos-Renyi control is **distinguishable**. -### The verdict depends on the extract, recurrence predicts it, and that is now the result +### The verdict depends on the extract; the saved scale ladder points to recurrence The single most important thing to know before writing any dependence sentence: **the same analysis, with the same instrument and no parameter changed, gives opposite answers on different extracts of one connectome.** On the sparse 1-hop `named` cut — an in-star, most of whose edges terminate on four seed cells and most of whose nodes receive no input — most cells come back composition-dominated. On the denser `taste_motor` cut, none does. -The scaling study (`flylab.analysis.scale.dependence_vs_scale`, T27) then says what the inversion tracks, and the answer is **not** size. Across seven extracts the composition-dominated verdict survives on exactly one — the in-star — while `scale_1k` has *fewer* nodes than `named` (1000 against 1126) and a mean degree of 22.9 against 1.21, and is already topology-dependent. At the 5k and 10k rungs the prediction is distinguishable from every null on the ladder. A degree-preserving rewire of an in-star is close to the identity, so a negative topology verdict there is weak evidence rather than a finding. Three consequences: +The saved scaling study (`flylab.analysis.scale.dependence_vs_scale`, T27) +contains seven extracts. The composition-dominated verdict occurs only on the +sparse `named` in-star; `scale_1k` has fewer nodes than `named` (1000 against +1126), a mean degree of 22.9 instead of 1.21, and is already topology-dependent. +This is consistent with recurrence tracking the difference, but it does not +isolate a causal feature: the extracts differ in more than node count. The +larger `scale_5k`–`scale_50k` inputs are absent from the current checkout, so +T27 cannot be independently rebuilt here. + +At `scale_5k` and `scale_10k`, the imidacloprid effect is distinguishable from +all three ranked edge/wiring nulls (`weight_permute`, +`rewire_degree_preserving`, `erdos_renyi`). The plain transmitter-label null is +indeterminate at `scale_10k`; it is an orthogonal check outside the ranked +necessary-information ladder. At `scale_25k` and `scale_50k`, the three ranked +edge/wiring nulls remain distinguishable, while the two transmitter-label +checks are not both distinguishable. The 25k and 50k runs use n = 20, where the +minimum p is 0.048 against α = 0.05, so those rejections are the smallest the +test can express and confirm rather than establish the pattern. 1. **Never state a dependence class without the extract it was measured on**, and quote `flylab.analysis.dependence.cut_census` (mean degree, share of edges onto seeds, share of nodes with any input, recurrence budget) beside it. -2. **"Most predictions do not need the connectome" is withdrawn**, along with "the topology-dependent set is exactly the chloride-channel blockers". What replaces them is a methodological claim: the dependence verdict is a joint property of the prediction and the extract, **recurrence rather than size predicts which way it goes**, and that is the transferable result. -3. **Do not quote the top of the ladder as strong evidence.** The 25k and 50k rungs run at n = 20, where the smallest attainable probability is 0.048 against α = 0.05; they confirm a verdict settled at better-powered rungs rather than establishing it. +2. **"Most predictions do not need the connectome" is withdrawn**, along with "the topology-dependent set is exactly the chloride-channel blockers". What replaces them is a methodological claim: the dependence verdict is a joint property of the prediction and the extract; the saved extracts associate the inversion more with recurrence than node count, but do not show that recurrence alone causes it. +3. **Do not quote the top of the ladder as strong evidence.** The 25k and 50k rungs run at n = 20. Raising n requires restoring the exact cuts and validating any faster degree-preserving sampler against the same graph invariants and permutation procedure. -The instrument itself is not in doubt here: `ladder_recovery` and `ladder_power` plant a recurrent loop of known strength in a synthetic graph of matched size, density and composition, recover it, and give an empirical false-positive rate on the unplanted control. The reversal is therefore about the graphs, not about power. +The planted-cycle checks provide limited evidence that the ladder recovers that one synthetic mechanism; they do not establish general power for biological endpoints. The `named` versus `taste_motor` landscape uses the same n = 1000 effort, so its contrast is conditional on those two extracts and settings. The T27 scale rungs use different, often smaller permutation counts; that trend is descriptive and cannot rule out power differences. ## Current headline comparisons @@ -117,7 +134,7 @@ Describe the landscape as: > "Topology-sensitive classifications are concentrated among perturbations affecting inhibitory signalling, with mechanism and concentration-specific exceptions." -The exact counts and classifications should be taken from papers/results.json, not copied by hand into durable project descriptions. That record has not been regenerated since the current round of revisions, so check any number against the shipped defaults before quoting it. +The exact counts and classifications should be taken from the generated `papers/results.json` and identified with the shipped defaults. T27 is an imported scale-study record; its 5k–50k input cuts are not in this checkout. ## Specification robustness diff --git a/docs/PAGES.md b/docs/PAGES.md index 0756c13..14edebb 100644 --- a/docs/PAGES.md +++ b/docs/PAGES.md @@ -26,6 +26,12 @@ The static application reaches the same core through flylab/browser/bridge.py. The user interface is shared. +The bench stores the primary run and experiment-design settings in the current +browser. **Export setup** writes a versioned JSON preset that can be imported +in another browser build; it contains settings, not assay results. At phone +widths, the FlyLab and Bench tab rows become a single grouped native panel +picker, synchronized with the active panel. + ## What "same core" means The browser and server builds share: @@ -177,13 +183,21 @@ Fast test suite: python -m pytest -q ~~~ -Longer end-to-end browser and static-build checks: +Longer end-to-end browser and static-build checks (the browser smoke test +requires Playwright and Chromium): ~~~bash -pytest -m slow +python -m pip install -e ".[dev,viz,browser-test]" +python -m playwright install chromium +pytest -q -m slow ~~~ -The ordinary tests workflow currently runs the fast suite. Slow checks should be run before tagged research releases and can also be placed in a scheduled or release-specific CI job. +The scheduled and manually dispatched slow CI job installs Chromium and runs +the static browser smoke test. It builds the Pages site, boots Pyodide in a +headless browser, runs the dashboard through the bridge, switches panels, and +blocks the chart scripts to check that their data tables remain available. +Pyodide and its packages still load from the configured runtime URLs in this +test; it does not verify a fully offline browser session. ## Performance diff --git a/docs/RESEARCH_MAP.md b/docs/RESEARCH_MAP.md index f58e92c..0331815 100644 --- a/docs/RESEARCH_MAP.md +++ b/docs/RESEARCH_MAP.md @@ -49,9 +49,9 @@ Public feathers (`body-annotations…`, `body-neurotransmitters…`, `connectome | 6d Specification robustness | 25 gain specifications x 7 conclusions + threshold grid | **done in v0.6**, but its topology rows were vacuous until **v0.6.1** | | 6e Global uncertainty + VOI | Sobol' budget, validated estimator, ranked experiments | **done in v0.6** | | **6f Substrate check** | repeat every dependence claim on a cut that is not an in-star | **done in v0.6.1 — and the verdict reversed** | -| **6i Scale ladder (new)** | the same profile across nested extracts of 1k-50k cells | **done in v0.6.1** — the verdict settles as topology-dependent, and recurrence rather than size predicts the inversion (T27) | -| 7 Expression coverage | adult motor-neuron receptor expression, or a declared bound | **open — 0.205, MN9 unresolved** | -| 7b Better power at the top of the ladder | the 25k/50k rungs run at n = 20 (resolution 0.048) | **open** — they confirm the settled verdict rather than establishing it | +| **6i Scale ladder (new)** | the same profile across nested extracts of 1k-50k cells | **saved result in v0.6.1** — T27; exact 5k-50k cuts are absent from this checkout, so those rungs cannot currently be rebuilt | +| 7 Expression evidence | map adult VNC motor-neuron labels to reprocessed counts; resolve MN9 separately | **open — GSE141807 now has VNC-wide gene-detection fractions; cell classes and MN9 remain unresolved** | +| 7b Better power at the top of the ladder | rerun exact 5k-50k cuts with higher permutation resolution | **gated** — cuts are missing here; profile exact edge swaps before raising n | | 8 Archival | Zenodo deposit; only then may "prospective" become "pre-registered" | **open** | ### Gate 1c is closed (former issue #2) @@ -86,7 +86,7 @@ H1 dual panel gap at 1 µM · H2 imidacloprid lowers the neighbourhood mean rate Every number here comes from `papers/results.json`. -1. **The dependence verdict is a joint property of the prediction and the extract, and recurrence predicts which way it goes.** Over 21 compounds × 4 concentrations at n = 1000 with Benjamini-Hochberg and a 1 % relative effect floor, the 1-hop `named` cut gives 11 topology-dependent, 42 composition-dominated and 31 no-effect; repeated unchanged on the denser `taste_motor` cut it gives 54 / 0 / 30. The scaling study over seven extracts (T27) shows the composition-dominated verdict surviving on exactly one of them — the in-star — while `scale_1k`, with *fewer* nodes than `named` and twenty times the mean degree, is already topology-dependent. **"Most predictions do not need the connectome" is withdrawn**, and so is "exactly the chloride-channel blockers". Report the extract's structure (T19, T27) with any dependence claim. +1. **The dependence verdict is a joint property of the prediction and the extract; in the saved scale study, the inversion is associated with recurrence.** Over 21 compounds × 4 concentrations at n = 1000 with Benjamini-Hochberg and a 1 % relative effect floor, the 1-hop `named` cut gives 11 topology-dependent, 42 composition-dominated and 31 no-effect; repeated unchanged on the denser `taste_motor` cut it gives 54 / 0 / 30. The saved scaling study over seven extracts (T27) shows the composition-dominated verdict surviving on exactly one of them — the in-star — while `scale_1k`, with *fewer* nodes than `named` and twenty times the mean degree, is already topology-dependent. This is consistent with recurrence tracking the change, but does not isolate a causal structural feature; the larger input cuts are absent from this checkout. **"Most predictions do not need the connectome" is withdrawn**, and so is "exactly the chloride-channel blockers". Report the extract's structure (T19, T27) with any dependence claim. 2. **The bitter-veto arm is a properly powered negative.** At n = 300 on the taste-motor cut, fipronil's veto ratio gives p = 0.42 / 0.88 / 0.20 / 0.24 with all |z| < 0.25, replacing the ten-shuffle hand-wave of v0.5. H6 is a pharmacology prediction, not a wiring prediction, and that is now measured rather than assumed. Imidacloprid's veto ratio there is **undefined** (MN9 silenced), not zero. 3. **The ordering information is in the mechanism rules — but the number that says so is nearly an identity.** Composition-only tracks the full model's ordering (ρ 0.988 at 1 µM), yet pharmacology-free pseudo-compounds reach a median of 0.847 and permuting compound labels leaves it identically 0.988; under a degree-corrected engine it falls to 0.666 (and -0.372 at 10 nM). Receptor-only orders the library backwards (ρ -0.506 to -0.046). Topology-only with a **direction-aware** generic rule reaches ρ 0.961, against 0.242 for the old depression-only rule — so "the connectome without the pharmacology is not a cheap substitute" is **withdrawn**. 4. **The suppression result is specification-dependent, and the old 25/25 was vacuous.** C1 is retained by 15 of 25 and reversed by all 10 monotone rules. The v0.6 topology rows were computed at 6 shuffles and under a specification context that never reached the null engine; recomputed at 100 shuffles with FDR across 100 structural tests, C3 holds 25/25 of which only 17 reach equivalence and 8 are indeterminate. Nicotinic buffering 18/25; Nav/AChE amplification 19/25. Retained by every specification: C2_rdl_disinhibition, C3_imidacloprid_topology_not_distinguishable, C4_fipronil_topology_exceeds, C7_map_bitter_veto. @@ -110,7 +110,7 @@ Every number here comes from `papers/results.json`. ## v0.7 (the next things, in order) -1. **Expression coverage.** 0.205 across the insect receptor keys, and motor neurons — MN9's class — are a confirmed gap. Either source it or declare the bound permanent. A data problem, not a modelling one. +1. **Expression evidence.** 0.205 is the fraction of graph-node × receptor-key pairs with a coarse class-level annotation across the two committed cuts, not a fraction of cells measured to express receptors. GEO GSE141807 is reprocessed into gene-detection fractions after published QC, but this is VNC-wide and has no joined cell-class labels. Determine whether adult motor-neuron annotations can be matched to those barcodes; the data do not identify MN9, whose proboscis motor-neuron arbors are in the brain subesophageal zone. Resolve MN9 with a class-matched brain/SEZ atlas or a targeted assay. 2. **The one allowed fit, now named by the model.** The engagement→gain transformation. Not a potency: at a saturating dose the model cannot see one. 3. **The free win.** `weight_threshold` needs no bench time: re-run against synapse-confidence strata. 4. **One live assay.** Climbing first; the PER baseline could not be sourced and is `null`. diff --git a/docs/REVIEW_ROUND2_LIBRARY_AUDIT.csv b/docs/REVIEW_ROUND2_LIBRARY_AUDIT.csv new file mode 100644 index 0000000..104f5fc --- /dev/null +++ b/docs/REVIEW_ROUND2_LIBRARY_AUDIT.csv @@ -0,0 +1,47 @@ +index,row_key,parameter,value_M,relation,species,source_has_identifier,current_source,candidate_identifier,support_verdict,recommended_addition_or_limitation,evidence_url +1,imidacloprid/insect_nAChR,EC50,2e-08,exact_compound_related_receptor,Ctenocephalides felis alpha1/alpha2 + Gallus gallus beta2 hybrid,True,"Dederer H, Werr M, Ilg T (2011) Insect Biochem Mol Biol 41(1):51-61. DOI 10.1016/j.ibmb.2010.09.012; PMID 20933086. Imidacloprid EC50 was 0.02 uM on cat-flea Cfalpha1/chicken Gg-beta2 hybrid receptors. The same paper also studied Drosophila alpha subunits with chicken beta2, but the 0.02 uM value is not a Drosophila receptor measurement. This unchanged library value is therefore an E2 related-receptor functional-engagement proxy, not binding occupancy.",10.1016/j.ibmb.2010.09.012; PMID 20933086,"direct, transferred","0.02 µM is on cat-flea alpha/chicken beta2 hybrid, not Drosophila. Retain E2 proxy; specify preparation.",https://pubmed.ncbi.nlm.nih.gov/20933086/ +2,imidacloprid/vertebrate_nAChR_a4b2,EC50,1e-05,class_extrapolation,mammalian (review-level contrast),False,"Tomizawa and Casida 2005, Annu Rev Pharmacol Toxicol 45:247 (imidacloprid selectivity for insect over vertebrate nAChR is >1000-fold; vertebrate potency in the micromolar order). Order-of-magnitude selectivity contrast, kept from FlyLab v0.4.",10.1146/annurev.pharmtox.45.120403.095930,class-order extrapolation,Review supports insect/mammal contrast; 10 µM is not a measured human alpha4beta2 EC50. Label illustrative.,https://doi.org/10.1146/annurev.pharmtox.45.120403.095930 +3,imidacloprid/insect_nAChR_native_dmel,Kd,2e-09,exact_compound_exact_receptor_exact_species,Drosophila melanogaster (native head membranes; nAChR population not resolved to subunits),True,"Tomizawa, Latli and Casida 1996, J Neurochem 67(4):1669-1676, doi:10.1046/j.1471-4159.1996.67041669.x, PMID 8858952 (""Head membranes of Drosophila melanogaster and Musca domestica have a single high-affinity binding site for [3H]IMI with KD values of 1-2 nM and Bmax values of 560-850 fmol/mg of protein""). This is a radioligand saturation Kd measured on the MODELLED organism, which is why it is the library's only binding_occupancy row. Two honest limits are recorded rather than smoothed over: (a) the publication states the 1-2 nM range for Drosophila and Musca jointly and does not resolve the Drosophila-only value, so FlyLab takes the conservative (weakest-affinity) end, 2.0e-9 M, and the tier stays literature_order; (b) the preparation is native head membranes and resolves NO subunit - the companion photoaffinity study (Tomizawa and Casida 1997, Neurosci Lett 237:61, PMID 9453215, doi:10.1016/s0304-3940(97)00811-2) labels a single 66 kDa ligand-binding polypeptide of the Drosophila nAChR without assigning it to a named Dalpha/Dbeta subunit - so the row sits on the preparation-resolved key insect_nAChR_native_dmel and makes no subunit claim. It does not replace the Dederer hybrid-receptor EC50 on the aggregate insect_nAChR key, which is a functional potency and stays where it was.",10.1046/j.1471-4159.1996.67041669.x; PMID 8858952,direct range,"Native-head-membrane Drosophila/Musca Kd 1-2 nM jointly; 2 nM is range endpoint, not Drosophila-specific or subunit-specific.",https://pubmed.ncbi.nlm.nih.gov/8858952/ +4,imidacloprid/insect_nAChR_beta1,Kd,8.3e-11,exact_compound_exact_receptor_other_species,"Myzus persicae (native beta1-containing nAChR, susceptible clone 4106A)",True,"Bass et al. 2011, BMC Neuroscience 12:51, doi:10.1186/1471-2202-12-51 ([3H]-imidacloprid saturation binding on whole-body membranes: the susceptible 4106A clone has two sites, Kd 0.083 +/- 0.0085 nM (Bmax 100 fmol/mg) and Kd 1.7 +/- 0.65 nM (Bmax 298 fmol/mg); the R81T mutation in the nAChR beta1 subunit abolishes the high-affinity site, leaving a single Kd 4.14 +/- 0.39 nM site in the resistant FRC clone). FlyLab uses the high-affinity Kd = 8.3e-11 M. This is a genuine equilibrium dissociation constant from a radioligand binding assay, but it was measured in an APHID, not in Drosophila. Under the v0.6.1 rule a binding constant licenses physical binding_occupancy only with relation exact_compound_exact_receptor_exact_species, so this row sits at evidence distance E1 and drives binding_engagement_proxy: the same Hill number, reported as an engagement proxy derived from a binding constant measured in another species, and carrying a provenance warning that names the gap. The transfer of an aphid beta1 affinity to the Drosophila Dbeta1 receptor is an assumption, not a measurement.",10.1186/1471-2202-12-51,"direct, transferred",0.083 nM aphid high-affinity binding site; beta1 mutation association does not make this Drosophila Dbeta1 affinity. Retain E1 proxy only.,https://doi.org/10.1186/1471-2202-12-51 +5,nitenpyram/insect_nAChR,EC50,5e-08,class_extrapolation,insect (review-level class statement),False,"Tomizawa and Casida 2005, Annu Rev Pharmacol Toxicol 45:247 (nitrocontaining neonicotinoids bind insect nAChR with low-nM to tens-of-nM affinity). Class-order value kept unchanged from FlyLab v0.4. TYPING NOTE (v0.6): the cited review mixes [3H]imidacloprid displacement affinities with functional potencies, so this row is typed as the weaker claim (EC50 / class_extrapolation) and drives functional_engagement only.",10.1146/annurev.pharmtox.45.120403.095930,class-order extrapolation,"Review discusses neonicotinoid binding classes, not nitenpyram functional EC50 50 nM. Do not call measured EC50.",https://doi.org/10.1146/annurev.pharmtox.45.120403.095930 +6,nitenpyram/vertebrate_nAChR_a4b2,EC50,2e-05,class_extrapolation,mammalian (review-level contrast),False,"Tomizawa and Casida 2005, Annu Rev Pharmacol Toxicol 45:247 (vertebrate nAChR potency of nitro-neonicotinoids is micromolar or weaker). Order-of-magnitude selectivity contrast, kept from FlyLab v0.4.",10.1146/annurev.pharmtox.45.120403.095930,class-order extrapolation,20 µM is not a measured human alpha4beta2 EC50 for nitenpyram. Keep illustrative only.,https://doi.org/10.1146/annurev.pharmtox.45.120403.095930 +7,nicotine/insect_nAChR,EC50,2e-07,exact_compound_related_receptor,Drosophila melanogaster Dalpha1 + Gallus gallus beta2 hybrid,False,"Ihara et al. 2018, Br J Pharmacol 175:1999 (Drosophila Dalpha1/chicken beta2 hybrid nAChR; nicotine active in the sub-micromolar order). Value kept unchanged from FlyLab v0.4.",10.1111/bph.13914; PMID 28616862,paper verified; value unconfirmed,Drosophila Dalpha1/chicken beta2 hybrid studied. Check figure/table for nicotine EC50 0.2 µM before claiming direct numerical support.,https://pubmed.ncbi.nlm.nih.gov/28616862/ +8,nicotine/vertebrate_nAChR_a4b2,EC50,1e-06,exact_compound_exact_receptor_other_species,Homo sapiens alpha4beta2 in Xenopus oocytes,False,"Chavez-Noriega et al. 1997, J Pharmacol Exp Ther 280:346 (human alpha4beta2 expressed in oocytes; nicotine EC50 in the low-micromolar order). Value kept unchanged from FlyLab v0.4.",PMID 8996215,paper verified; value unconfirmed,Human alpha4beta2 oocyte pharmacology exists; nicotine 1 µM requires primary table check and stoichiometry context.,https://pubmed.ncbi.nlm.nih.gov/8996215/ +9,diazepam/vertebrate_GABA_A,EC50,1e-08,exact_compound_exact_receptor_other_species,mammalian alpha1beta2gamma2 GABA-A,False,"Sigel and Ernst 2018, Trends Pharmacol Sci 39:659 (benzodiazepine site of alpha1beta2gamma2 GABA-A; diazepam potentiation in the 10 nM order). Value kept unchanged from FlyLab v0.4. TYPING NOTE (v0.6): a genuine benzodiazepine-site Kd exists in the radioligand literature ([3H]flunitrazepam displacement), but the value used here comes from a POTENTIATION measurement, so the row is typed EC50 and, the preparation being a mammalian (non-Drosophila) receptor at evidence distance E1, reports functional_engagement_proxy, not binding occupancy. Upgrading it to Kd would require citing the binding study instead.",10.1016/j.tips.2018.03.006,review; value unverified,"Review of benzodiazepine binding sites is not a direct 10 nM diazepam potentiation assay. Locate primary study, state GABA co-application and assay conditions.",https://doi.org/10.1016/j.tips.2018.03.006 +10,fipronil/vertebrate_GABA_A,IC50,1.1e-06,exact_compound_exact_receptor_other_species,Homo sapiens recombinant GABA-A ([3H]EBOB displacement),False,"Ratra and Casida 2001, Toxicol Lett 122:215 ([3H]EBOB displacement, human recombinant GABA-A): vertebrate average IC50 for fipronil is 1103 nM = 1.1e-6 M. CORRECTED in v0.5: FlyLab v0.4 carried 1.0e-05 M while citing this same paper, so the library contradicted its own citation; the value is now the cited one. SUBUNIT CAVEAT: the human beta3 HOMOpentamer is far more sensitive, 0.5-2.4 nM (Ratra, Kamita and Casida 2001, Toxicol Appl Pharmacol 172:233), so the 'vertebrate-safe' reading of this row applies to native heteromeric receptors only, not to beta3 homomers.",10.1021/tx980157t; PMID 9860498,numeric direct; citation/species wrong,"1103 nM comes from Hainzl et al. 1998, multi-vertebrate [3H]EBOB displacement average. Replace Ratra attribution and human-recombinant label.",https://pubmed.ncbi.nlm.nih.gov/9860498/ +11,fipronil/insect_RDL,IC50,3e-08,exact_compound_exact_receptor_exact_species,Drosophila melanogaster,False,"Hosie et al. 1995, Br J Pharmacol 115:909 (fipronil block of dieldrin-sensitive Drosophila GABA receptors, tens-of-nM order). SPREAD NOTE: the literature spans ~1500-fold for this one row - 2.4-12 nM on house-fly head membranes ([3H]EBOB displacement; Cole et al. 1993, Pestic Biochem Physiol 46:47 and Ratra and Casida 2001, Toxicol Lett 122:215) versus 0.3-3 uM on the predominant in vivo splice/edit variant of Drosophila RDL in Xenopus oocytes (Lees et al. 2014, PLoS ONE 9:e97468). Library keeps the potent end (unchanged from FlyLab v0.4); it is a defensible midpoint-to-potent choice, not a measured constant.",10.1111/j.1476-5381.1995.tb15896.x; PMID 7582519,assay direct; value unconfirmed,"Hosie tests Drosophila RDL fipronil block, but 30 nM is not established by checked abstract; predominant RDL variant is ~0.63 µM in Lees 2014. Specify variant/assay or use Lees value.",https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0097468 +12,clothianidin/insect_nAChR,EC50,3e-08,exact_compound_related_receptor,Periplaneta americana native neuronal nAChR,False,"Thany 2009, Neurotoxicology 30:1045 (clothianidin is an agonist at cockroach nAChRs from about 30 nM); Tomizawa and Casida 2005, Annu Rev Pharmacol Toxicol 45:247 for low-nM insect binding.",10.1016/j.neuro.2009.06.013; PMID 19583978,paper verified; value unconfirmed,Cockroach synaptic/extrasynaptic receptor agonism is supported; 30 nM exact EC50 needs full assay verification. Related-receptor transfer explicit.,https://pubmed.ncbi.nlm.nih.gov/19583978/ +13,clothianidin/vertebrate_nAChR_a7,EC50,0.00074,exact_compound_exact_receptor_other_species,Rattus norvegicus alpha7,False,"Cartereau et al. 2018, Br J Pharmacol 175 (rat alpha7; clothianidin partial agonist, EC50 0.74 mM, vs acetylcholine 0.16 mM).",10.1111/bph.14018; PMID 28853147,direct,Rat alpha7 oocyte clothianidin EC50 ~0.74 mM supported; retain species and partial agonism.,https://pubmed.ncbi.nlm.nih.gov/28853147/ +14,clothianidin/vertebrate_nAChR_a4b2,EC50,0.0001,class_extrapolation,mammalian (inferred from the alpha7 data),False,"Order-of-magnitude contrast from the mammalian alpha7 data of Cartereau et al. 2018, Br J Pharmacol 175; no direct alpha4beta2 EC50 was located, so this row is a high-micromolar class order, not a measured EC50.",10.1111/bph.14018; PMID 28853147,unsupported cross-receptor extrapolation,Alpha7 data cannot validate 0.1 mM alpha4beta2 EC50. Label a synthetic assumption or remove numeric row pending alpha4beta2 assay.,https://pubmed.ncbi.nlm.nih.gov/28853147/ +15,thiamethoxam/insect_nAChR,EC50,0.0001,exact_compound_related_receptor,Periplaneta americana neurons and cercal afferent/giant interneuron synapses,False,"Thany 2011, Neuropharmacology 60:587 (thiamethoxam is a poor agonist on isolated cockroach neuronal cell bodies but a full agonist at cercal afferent/giant interneuron synapses); Nauen et al. 2003, Pestic Biochem Physiol 76:55 (thiamethoxam is a precursor converted to clothianidin; binding up to 10000-fold weaker than other neonicotinoids).",10.1016/j.neuropharm.2010.12.008; PMID 21172360,paper verified; value unconfirmed,TMX is poor agonist on isolated cockroach cell bodies but full agonist at synapse; 0.1 mM depends on preparation and needs figure check.,https://pubmed.ncbi.nlm.nih.gov/21172360/ +16,acetamiprid/insect_nAChR,EC50,5e-08,class_extrapolation,insect (review-level class statement),False,"Tomizawa and Casida 2005, Annu Rev Pharmacol Toxicol 45:247 (cyanoamidine neonicotinoids bind insect nAChR with low-nM to tens-of-nM affinity). TYPING NOTE (v0.6): review-level affinity statement; typed as the weaker claim (EC50 / class_extrapolation, evidence distance E3), so it drives functional_engagement_proxy only.",10.1146/annurev.pharmtox.45.120403.095930,class-order extrapolation,Review class trend does not establish acetamiprid functional EC50 50 nM. Distinguish binding from activation.,https://doi.org/10.1146/annurev.pharmtox.45.120403.095930 +17,acetamiprid/vertebrate_nAChR_a7,EC50,0.00073,exact_compound_exact_receptor_other_species,Rattus norvegicus alpha7,False,"Cartereau et al. 2018, Br J Pharmacol 175 (rat alpha7; acetamiprid EC50 0.73 mM).",10.1111/bph.14018; PMID 28853147,direct,Rat alpha7 oocyte acetamiprid EC50 ~0.73 mM supported; retain species and partial agonism.,https://pubmed.ncbi.nlm.nih.gov/28853147/ +18,acetamiprid/vertebrate_nAChR_a4b2,EC50,0.0001,class_extrapolation,mammalian (inferred from the alpha7 data),False,"Order-of-magnitude contrast from the mammalian alpha7 data of Cartereau et al. 2018, Br J Pharmacol 175; not a measured alpha4beta2 EC50.",10.1111/bph.14018; PMID 28853147,unsupported cross-receptor extrapolation,Alpha7 data cannot validate 0.1 mM alpha4beta2 EC50. Label synthetic assumption or remove pending assay.,https://pubmed.ncbi.nlm.nih.gov/28853147/ +19,sulfoxaflor/insect_nAChR,EC50,0.00015,exact_compound_related_receptor,Periplaneta americana DUM neurons (alpha-bungarotoxin-insensitive nAChR),False,"Watson et al. 2011, Pestic Biochem Physiol 101:32 (sulfoxaflor fully activates alpha-bungarotoxin-insensitive nAChRs of cockroach DUM neurons, EC50 about 154 uM); Sparks et al. 2013, Pestic Biochem Physiol 107:1 (mode of action review).",10.1016/j.ibmb.2011.01.009; PMID 21296156,source citation wrong; number unverified,"Matching Watson 2011 primary paper is IBMB 41:432-439, not Pestic Biochem Physiol 101:32. It supports Dalpha2/chicken beta2 efficacy but 154 µM needs table confirmation.",https://pubmed.ncbi.nlm.nih.gov/21296156/ +20,sulfoxaflor/vertebrate_nAChR_a4b2,EC50,0.001,class_extrapolation,mammalian (review-level contrast),False,"Sparks et al. 2013, Pestic Biochem Physiol 107:1 (low mammalian nAChR activity reported for sulfoximines). Order-of-magnitude contrast, not a measured EC50.",10.1016/j.pestbp.2013.05.014; PMID 25149228,review/class extrapolation,Sparks 2013 review does not establish sulfoxaflor alpha4beta2 EC50 1 mM. Label synthetic contrast.,https://pubmed.ncbi.nlm.nih.gov/25149228/ +21,spinosad/insect_nAChR,EC50,5e-06,exact_compound_related_receptor,Drosophila melanogaster Dalpha6 / heterologous insect alpha6-like receptors,False,"Perry et al. 2015, Insect Biochem Mol Biol 64:106 (Drosophila Dalpha6 is required for spinosad action in vivo); heterologous alpha6 nAChR studies report allosteric activation in the low-micromolar order (EC50 about 5 uM). AGGREGATE ROW (v0.6): the measurement behind this number is alpha6-specific and is recorded subunit-resolved in insect_nAChR_alpha6. The same value is repeated here on the documented aggregate key (no new number was invented) so the legacy g_ach gain rule still fires; read as an aggregate it overstates the receptor population the value applies to.",10.1016/j.ibmb.2015.01.017; PMID 25747008,target supported; number and direction unsupported,Perry 2015 in-vivo rescue paper lacks heterologous receptor EC50. Later Drosophila work reports alpha6 antagonism. Remove 5 µM agonist value pending direct assay.,https://pubmed.ncbi.nlm.nih.gov/25747008/ +22,spinosad/insect_nAChR_alpha6,EC50,5e-06,exact_compound_related_receptor,Drosophila melanogaster Dalpha6 (in vivo requirement) / heterologous insect alpha6-like receptors (the exact preparation behind the 5 uM order is not pinned),False,"Perry et al. 2015, Insect Biochem Mol Biol 64:106 (Drosophila Dalpha6 is required for spinosad action in vivo; Dalpha6 knockouts are highly resistant); heterologous alpha6 nAChR studies report allosteric activation in the low-micromolar order (EC50 about 5 uM). SUBUNIT-RESOLVED ROW (v0.6): this is the assay the spinosad number actually came from. The identical value also appears on the aggregate insect_nAChR key so the legacy gain rules keep working; no new number was invented for the split.",10.1016/j.ibmb.2015.01.017; PMID 25747008,target supported; number and direction unsupported,Same as aggregate row: Dalpha6 target genetics do not measure a 5 µM EC50; check direction against antagonist data.,https://pubmed.ncbi.nlm.nih.gov/35191376/ +23,acetylcholine/insect_nAChR,EC50,1e-05,exact_compound_related_receptor,Drosophila melanogaster Dalpha1 + Gallus gallus beta2 hybrid (and Dalpha5 homomers),False,"Ihara et al. 2018, Br J Pharmacol 175:1999 (Drosophila Dalpha1/chicken beta2 hybrids; acetylcholine reference responses in the low-micromolar range). Drosophila Dalpha5 homomers report an acetylcholine EC50 of about 8.8 uM.",10.1111/bph.13914; PMID 28616862,paper verified; value unconfirmed,Hybrid-receptor acetylcholine reference response exists; 10 µM exact EC50 and Dalpha5 homomer 8.8 µM need primary table check.,https://pubmed.ncbi.nlm.nih.gov/28616862/ +24,acetylcholine/vertebrate_nAChR_a4b2,EC50,1e-05,exact_compound_exact_receptor_other_species,Homo sapiens alpha4beta2 in Xenopus oocytes,False,"Chavez-Noriega et al. 1997, J Pharmacol Exp Ther 280:346 (human alpha4beta2 in oocytes; acetylcholine EC50 in the low-micromolar range).",PMID 8996215,paper verified; value unconfirmed,Human alpha4beta2 ACh assay exists; 10 µM is assay/stoichiometry dependent and needs primary table check.,https://pubmed.ncbi.nlm.nih.gov/8996215/ +25,acetylcholine/vertebrate_nAChR_a7,EC50,0.00016,exact_compound_exact_receptor_other_species,Rattus norvegicus alpha7,False,"Cartereau et al. 2018, Br J Pharmacol 175 (rat alpha7; acetylcholine EC50 0.16 mM).",10.1111/bph.14018; PMID 28853147,direct,Rat alpha7 ACh EC50 ~0.16 mM in same primary study.,https://pubmed.ncbi.nlm.nih.gov/28853147/ +26,gaba/insect_RDL,EC50,5e-05,exact_compound_exact_receptor_exact_species,Drosophila melanogaster RDLbd in Xenopus oocytes,False,"Lees et al. 2014, PLoS ONE 9:e97468 (RDLbd and RDLbd I/V in Xenopus oocytes; GABA pEC50 4.2-4.3, i.e. about 50 uM).",10.1371/journal.pone.0097468; PMID 24823815,direct,"GABA pEC50 4.34 on RDLbd I/V = 46 µM, matching 50 µM rounding.",https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0097468 +27,gaba/vertebrate_GABA_A,EC50,1e-05,exact_compound_exact_receptor_other_species,mammalian alpha1beta2gamma2 GABA-A,False,"Sigel and Steinmann 2012, J Biol Chem 287:40224 (GABA-A receptors; GABA EC50 in the 5-30 uM range for alpha1beta2gamma2).",10.1074/jbc.R112.386664; PMID 23038269,review/range,Sigel-Steinmann is review. 10 µM within stated range but not a single primary alpha1beta2gamma2 result. Cite primary assay for exact parameter.,https://pubmed.ncbi.nlm.nih.gov/23038269/ +28,picrotoxin/insect_RDL,IC50,1e-06,exact_compound_exact_receptor_exact_species,Drosophila melanogaster,False,"Hosie et al. 1995, Br J Pharmacol 115:909 (picrotoxinin block of Drosophila GABA receptors, micromolar order); RDL oocyte IC50 varies up to about 17-fold with expression conditions (PLoS ONE 2021, RDL expression-condition study).",10.1111/j.1476-5381.1995.tb15896.x; PMID 7582519; 10.1371/journal.pone.0097468,assay direct; compound identity caveat,"Picrotoxin mixture versus picrotoxinin alone should be distinguished. Lees RDLbd I/V pIC50 6.44 = 0.36 µM, within order of 1 µM.",https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0097468 +29,picrotoxin/vertebrate_GABA_A,IC50,1e-06,exact_compound_exact_receptor_other_species,mammalian GABA-A,False,"Ratra and Casida 2001, Toxicol Lett 122:215 (picrotoxinin blocks mammalian GABA-A at about 1 uM, i.e. no useful insect selectivity).",10.1016/S0378-4274(01)00366-6; PMID 11489356,citation not direct,"Ratra-Casida is [3H]EBOB insecticide/subunit study, not demonstrated 1 µM picrotoxinin functional block. Find direct GABA-A assay.",https://pubmed.ncbi.nlm.nih.gov/11489356/ +30,picrotoxin/vertebrate_GlyR,IC50,1e-05,class_extrapolation,mammalian homomeric glycine receptor,False,"Ratra and Casida 2001, Toxicol Lett 122:215 and standard glycine receptor pharmacology (picrotoxin blocks homomeric GlyR in the micromolar order). Order-of-magnitude value.",10.1016/S0378-4274(01)00366-6; PMID 11489356,source-receptor mismatch,"Ratra-Casida examines GABA receptors, not GlyR. Find a glycine-receptor picrotoxin primary assay; 10 µM currently unsupported.",https://pubmed.ncbi.nlm.nih.gov/11489356/ +31,dieldrin/insect_RDL,IC50,1e-07,exact_compound_exact_receptor_exact_species,Drosophila melanogaster,False,"ffrench-Constant et al. 1993, Nature 363:449 (Rdl A301S confers cyclodiene resistance); Hosie et al. 1995, Br J Pharmacol 115:909 (cyclodiene block of dieldrin-sensitive Drosophila GABA receptors in the 0.1 uM order).",10.1038/363449a0; 10.1111/j.1476-5381.1995.tb15896.x,qualitative resistance; value unverified,ffrench-Constant proves RDL resistance mutation; Hosie paper measures fipronil. Neither checked source directly establishes dieldrin IC50 0.1 µM. Find primary dose-response.,https://doi.org/10.1038/363449a0 +32,dieldrin/vertebrate_GABA_A,IC50,1e-06,exact_compound_exact_receptor_other_species,mammalian GABA-A,False,"Ratra and Casida 2001, Toxicol Lett 122:215 (cyclodienes block mammalian GABA-A at sub-micromolar to micromolar; the insect selectivity margin is small).",10.1016/S0378-4274(01)00366-6; PMID 11489356,class-order extrapolation,Ratra-Casida supports GABA subunit dependence but exact dieldrin 1 µM mammalian IC50 was not verified.,https://pubmed.ncbi.nlm.nih.gov/11489356/ +33,ivermectin/insect_GluCl,EC50,2e-08,exact_compound_related_receptor,invertebrate GluCl (Drosophila genetics plus heterologous invertebrate GluCl),False,"Kane et al. 2000, PNAS 97:13949 (drug-resistant Drosophila show GluCl is the avermectin target); invertebrate GluCl activation is reported in the 0.1-30 nM order.",10.1073/pnas.240464697; PMID 11095718,target supported; value unverified,"Kane 2000 supports Drosophila GluCl target/resistance, not directly checked 20 nM functional EC50. Need heterologous Dmel GluCl dose-response.",https://doi.org/10.1073/pnas.240464697 +34,ivermectin/insect_RDL,EC50,1e-06,exact_compound_exact_receptor_exact_species,Drosophila melanogaster RDL (predominant in vivo variant) in Xenopus oocytes,False,"Lees et al. 2014, PLoS ONE 9:e97468 (ivermectin activates and potentiates the predominant in vivo RDL variant at sub-micromolar to micromolar).",10.1371/journal.pone.0097468; PMID 24823815,contradicted mechanism,"Lees reports ivermectin inhibits GABA responses, no current alone at 10 µM; RDLbd I/V pIC50 6.33 (~0.47 µM). Change positive_modulator/EC50 interpretation after model audit.",https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0097468 +35,ivermectin/vertebrate_GlyR,EC50,1e-06,exact_compound_exact_receptor_other_species,rat/human alpha1 glycine receptor,False,"Shan et al. 2001, J Biol Chem 276:12556 (ivermectin is an unconventional agonist of the alpha1 glycine receptor, sub-micromolar to micromolar).",10.1074/jbc.M011264200; PMID 11278873,direct mechanism; value unconfirmed,Shan shows human alpha1 GlyR potentiation at 0.03 µM and direct activation above 0.03 µM. 1 µM lumping needs assay-specific curve.,https://pubmed.ncbi.nlm.nih.gov/11278873/ +36,ivermectin/vertebrate_GABA_A,EC50,1e-05,exact_compound_exact_receptor_other_species,Mus musculus GABA-A,False,"Dawson et al. 2000, J Pharmacol Exp Ther 295:1051 (avermectin effects in mice are mediated through GABA-A; micromolar potentiation).",PMID 11082440,paper verified; value unconfirmed,"Dawson examines avermectin analogs, mouse seizures/GABA-A; not necessarily ivermectin 10 µM functional EC50. Identify compound and in-vitro figure.",https://pubmed.ncbi.nlm.nih.gov/11082440/ +37,chlorpyrifos_oxon/insect_AChE,IC50,1e-08,exact_compound_exact_receptor_exact_species,Drosophila melanogaster AChE (Ace),False,"Fournier and Mutero 1994, Comp Biochem Physiol C 108:19 (Drosophila AChE is inhibited by oxon organophosphates in the low-nM order and is more sensitive than mammalian AChE).",10.1016/1367-8280(94)90084-1,review; number unsupported,"Fournier-Mutero 1994 is resistance review, not chlorpyrifos-oxon 10 nM Drosophila assay. Need primary enzyme IC50 and exposure time.",https://hero.epa.gov/reference/4491250/ +38,chlorpyrifos_oxon/vertebrate_AChE,IC50,3e-07,exact_compound_exact_receptor_other_species,Homo sapiens erythrocyte AChE in donor blood,False,"Archives of Toxicology 2020, interspecies AChE inhibition study in electric eel and human donor blood (chlorpyrifos-oxon IC50 about 0.3 uM in human blood).",10.1007/s00204-020-02927-8; PMID 33037899,"direct, assay-specific","Kasteel 2020 Table 2 gives CPO IC50 0.27 µM in human donor blood, supporting 0.3 µM rounded; blood, not isolated human enzyme.",https://pubmed.ncbi.nlm.nih.gov/33037899/ +39,deltamethrin/insect_Nav,EC50,4.3e-08,exact_compound_exact_receptor_exact_species,Drosophila melanogaster para in Xenopus oocytes,False,"Burton et al. 2011, Insect Biochem Mol Biol 41:723 (Drosophila para in Xenopus oocytes; tail-current EC50 0.043 uM deltamethrin, 0.40 uM permethrin, 65 uM DDT).",10.1016/j.ibmb.2011.05.004; PMID 21640822,direct,Burton Drosophila para tail-current EC50 deltamethrin 0.043 µM.,https://pubmed.ncbi.nlm.nih.gov/21640822/ +40,deltamethrin/vertebrate_Nav1_x,EC50,5e-06,exact_compound_exact_receptor_other_species,Rattus norvegicus Nav1.6,False,"Tan and Soderlund 2010, Neurotoxicology 31:16 (modification of rat Nav1.6 by deltamethrin requires micromolar concentrations).",10.1016/j.taap.2010.07.001; PMID 20624410,citation wrong; value unverified,"Tan-Soderlund 2010 rat Nav1.6 paper is TAAP 247:229-237, not Neurotoxicology 31:16. Verify 5 µM dose-response and state-dependent protocol.",https://pmc.ncbi.nlm.nih.gov/articles/PMC2929565/ +41,permethrin/insect_Nav,EC50,4e-07,exact_compound_exact_receptor_exact_species,Drosophila melanogaster para in Xenopus oocytes,False,"Burton et al. 2011, Insect Biochem Mol Biol 41:723 (Drosophila para tail-current EC50 0.40 uM for permethrin).",10.1016/j.ibmb.2011.05.004; PMID 21640822,direct,Burton Drosophila para tail-current EC50 permethrin 0.40 µM.,https://pubmed.ncbi.nlm.nih.gov/21640822/ +42,permethrin/vertebrate_Nav1_x,EC50,4e-05,class_extrapolation,Rattus norvegicus brain type IIA sodium channels (fold-difference statement),False,"Smith and Soderlund 1998, Neurotoxicology 19:823 (permethrin is more than 100-fold less potent on rat brain type IIA sodium channels than on insect para).",PMID 9863771,wrong compound,"Smith-Soderlund 1998 article measures cypermethrin on rat brain IIa, not permethrin; cannot support permethrin 40 µM. Find direct permethrin mammalian Nav assay.",https://pubmed.ncbi.nlm.nih.gov/9863771/ +43,ddt/insect_Nav,EC50,6.5e-05,exact_compound_exact_receptor_exact_species,Drosophila melanogaster para in Xenopus oocytes,False,"Burton et al. 2011, Insect Biochem Mol Biol 41:723 (Drosophila para tail-current EC50 65 uM for DDT, with only about 7 percent maximum channel modification).",10.1016/j.ibmb.2011.05.004; PMID 21640822,direct with low efficacy,Burton Drosophila para tail-current EC50 DDT 65 µM but maximum modification only 7%; model efficacy must preserve this weak maximum.,https://pubmed.ncbi.nlm.nih.gov/21640822/ +44,ddt/vertebrate_Nav1_x,EC50,0.001,class_extrapolation,mammalian sodium channels (review-level contrast),False,"Vijverberg and van den Bercken 1990, Crit Rev Toxicol 21:105 (DDT and pyrethroids modify mammalian sodium channels only at much higher concentrations than insect channels). Order-of-magnitude contrast.",no verified DOI/PMID,source mismatch,"Vijverberg-van den Bercken 1990 reviews pyrethroids, not a 1 mM DDT Nav1.x EC50. Find primary DDT mammalian assay or mark illustrative.",https://research-portal.uu.nl/en/publications/neurotoxicological-effects-and-the-mode-of-action-of-pyrethroid-i/ +45,chlordimeform/insect_OctR,EC50,3e-05,exact_compound_related_receptor,insect octopamine-sensitive adenylate cyclase preparation (octopamine-2 receptor),False,"Nathanson 1985, PNAS 82:599 (octopamine-sensitive adenylate cyclase; formamidines are potent and selective octopamine-2 receptor agonists). Reported Ka order: about 3e-5 M for chlordimeform and about 2e-6 M for its N-demethyl metabolite. TYPING NOTE (v0.6): the source reports an activation constant (Ka) for adenylate-cyclase stimulation, i.e. a functional potency. It is typed EC50 and drives functional_engagement; it is not a binding constant.",10.1073/pnas.82.2.599; PMID 2982165,direct functional; species transfer,"Nathanson studies firefly light-organ octopamine-sensitive adenylate cyclase; 30 µM parent chlordimeform Ka is functional estimate, not Drosophila receptor EC50.",https://pubmed.ncbi.nlm.nih.gov/2982165/ +46,caffeine/vertebrate_GABA_A,IC50,0.001,class_extrapolation,mammalian GABA-A (review-level statement),False,"Fredholm et al. 1999, Pharmacol Rev 51:83 (caffeine antagonises GABA-A only at millimolar concentrations, far above the adenosine-receptor range).",PMID 10049999; DOI 10.1016/S0031-6997(24)01396-6,review; number unverified,Fredholm 1999 caffeine review cannot alone establish 1 mM GABA-A IC50. Find direct receptor assay or mark illustrative.,https://pubmed.ncbi.nlm.nih.gov/10049999/ diff --git a/docs/REVIEW_ROUND2_LIBRARY_AUDIT.md b/docs/REVIEW_ROUND2_LIBRARY_AUDIT.md new file mode 100644 index 0000000..90c17d3 --- /dev/null +++ b/docs/REVIEW_ROUND2_LIBRARY_AUDIT.md @@ -0,0 +1,17 @@ +# FlyLab numeric library provenance audit (round 2) + +Scope: 46 non-null `value_M` receptor rows in the baseline `source/flylab/pharm/library.yaml`, read on 23 September 2026. The accompanying CSV records each row's baseline source, a verified or candidate identifier, evidence type, and a specific action. Seven values were subsequently withheld and two fipronil rows reclassified. The audit separates a missing DOI/PMID string from insufficient support for a numeric constant. A source's existence or a matching DOI does not establish its assay result. + +**Audit baseline:** The CSV preserves the original 46-row findings for traceability. The current library and generated T10 contain the subsequent corrections. + +## Immediate corrections before mechanistic claims + +1. **Spinosad (`insect_nAChR` and `insect_nAChR_alpha6`).** The cited [Perry et al. 2015](https://pubmed.ncbi.nlm.nih.gov/25747008/) establishes Dα6 involvement through in-vivo rescue, but says robust heterologous expression was absent. It does not substantiate a 5 µM receptor EC50. A later [Drosophila study](https://pubmed.ncbi.nlm.nih.gov/35191376/) reports low-dose spinosad antagonizing α6-mediated cholinergic response, so the agonist direction also needs reconsideration. Do not simply append Perry's DOI to these rows. +2. **Ivermectin (`insect_RDL`).** The cited [Lees et al. 2014](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0097468) directly contradicts `positive_modulator`: ivermectin suppressed GABA-evoked RDL currents, with no current when 10 µM ivermectin was applied alone. For RDLbd I/V it reports pIC50 6.33 (about 0.47 µM), an inhibitory parameter. The library's 1 µM *EC50* must be reviewed in the model as well as the source field. +3. **Fipronil (`vertebrate_GABA_A`).** The 1103 nM number is in [Hainzl, Cole and Casida 1998](https://pubmed.ncbi.nlm.nih.gov/9860498/), averaged over several vertebrates by radioligand displacement. The current source attributes it to a human recombinant receptor in Ratra and Casida 2001. The number can be retained only with the proper assay and species scope. +4. **Permethrin (`vertebrate_Nav1_x`).** The cited [Smith and Soderlund 1998](https://pubmed.ncbi.nlm.nih.gov/9863771/) concerns *cypermethrin*, not permethrin. It does not support the 40 µM permethrin row. +5. **DDT (`insect_Nav`).** [Burton et al. 2011](https://pubmed.ncbi.nlm.nih.gov/21640822/) directly supports 65 µM tail-current EC50 but reports only 7% maximum channel modification. This low efficacy must accompany any resulting mechanistic or selectivity claim. + +Other specific corrections: the Watson 2011 sulfoxaflor paper is [*Insect Biochem. Mol. Biol.* 41:432–439](https://pubmed.ncbi.nlm.nih.gov/21296156/), not *Pestic. Biochem. Physiol.* 101:32; 154 µM remains unconfirmed. The Tan and Soderlund 2010 rat Nav1.6 study is [*Toxicol. Appl. Pharmacol.* 247:229–237](https://pmc.ncbi.nlm.nih.gov/articles/PMC2929565/), not *Neurotoxicology* 31:16; its exact 5 µM model value remains unconfirmed. [Fournier and Mutero 1994](https://hero.epa.gov/reference/4491250/) is a resistance review, not a primary 10 nM chlorpyrifos-oxon assay. [Kasteel et al. 2020](https://pubmed.ncbi.nlm.nih.gov/33037899/) does directly support the *human-blood* chlorpyrifos-oxon IC50 of 0.27 µM (rounded to 0.3 µM in the library). + +The 43 numeric rows without literal DOI/PMID in their source strings should receive an identifier only after the source and its assay are matched. Class-order placeholder and cross-receptor rows should be labeled as model assumptions; identifier absence alone is not reason for E4 demotion. The CSV's “value unconfirmed” means the paper identity or abstract was checked but the exact table or figure was not, so it is an audit lead rather than a claim of falsity. diff --git a/docs/REVIEW_ROUND2_REFERENCE_AUDIT.md b/docs/REVIEW_ROUND2_REFERENCE_AUDIT.md new file mode 100644 index 0000000..23b772b --- /dev/null +++ b/docs/REVIEW_ROUND2_REFERENCE_AUDIT.md @@ -0,0 +1,19 @@ +# FlyLab reference audit, round 2 + +**Scope (23 September 2026).** I checked all 27 numbered entries in `FlyLab_manuscript_review_complete.docx` and all 76 numbered entries in `source/papers/IJRC_FlyLab_draft.md.in` (103 entry occurrences). The main bibliography has 25 DOI entries and two web entries. The source bibliography has 71 DOI entries and five without DOIs. Across them, 72 distinct DOIs resolved through [Crossref's registered-work API](https://api.crossref.org/works/10.1016%2Fj.cell.2026.08.015); registered titles, first authors, publication years, and volume/issue/pages or article numbers match the cited records. Apparent low title similarity was caused by HTML markup, truncation in Crossref, or punctuation/subtitles; no DOI pointed to a different paper. All 48 distinct listed PMIDs resolved through [NIH PubMed](https://pubmed.ncbi.nlm.nih.gov/) to the cited articles. Bibliographic existence is **not** proof that each paper supports a numerical library value or every scientific inference. + +| Location | Finding and precise action | Primary evidence | +| --- | --- | --- | +| Main [5]; source [6] | The round-2 review's uncertainty about Berg *et al.* is resolved. **Keep**: S. Berg *et al.*, “Sexual dimorphism in the complete *Drosophila* male central nervous system connectome,” *Cell*, vol. 189, no. 18, pp. 5504–5526.e15, 2026. DOI 10.1016/j.cell.2026.08.015. First author is Stuart Berg. | [PubMed PMID 42691995](https://pubmed.ncbi.nlm.nih.gov/42691995/); [Cell article](https://doi.org/10.1016/j.cell.2026.08.015) | +| Main [27]; source [76] and main §3.6 | Tomizawa *et al.* is valid, and the abstract reports **1–2 nM** high-affinity radiolabeled imidacloprid binding in *Drosophila* and *Musca* **head membranes**. If the manuscript uses a fixed 2 nM model input, state that it takes the upper end of the reported range. Avoid implying a precise *Drosophila*-only or isolated-receptor measurement. | [PubMed PMID 8858952](https://pubmed.ncbi.nlm.nih.gov/8858952/) | +| Source §3.7/licence statement (line 196) | Correct the joint licence assertion. MaleCNS says CC-BY; FlyWire's current public-release guidelines say **CC BY-NC 4.0**. Suggested wording: “MaleCNS v1.0 is released under CC-BY; FlyWire public-release data are licensed CC BY-NC 4.0. Their respective attribution and use terms apply.” | [MaleCNS download](https://male-cns.janelia.org/download/); [FlyWire citation guidelines](https://join.flywire.ai/guidelines) | +| Source bibliography [46] | The DOI, authors, year, volume, and article 182 are valid. The publisher's title spells “rice stem **borrers**” (a publisher typo); the bibliography silently corrects it to “borers.” For an exact-title policy, reproduce “borrers,” optionally with `[sic]`. Scientific meaning is unaffected. | [Oxford article](https://academic.oup.com/jinsectscience/article/10/1/182/887304) | +| Supplementary source inventory, “Original [54]” | This **extra** inventory entry still truncates the Joußen title after “*Nicotiana tabacum*,” although the numbered source/supplement bibliography [54] now contains the full title. Replace its title with: “Metabolism of imidacloprid and DDT by P450 CYP6G1 expressed in cell cultures of *Nicotiana tabacum* suggests detoxification of these insecticides in *Cyp6g1*-overexpressing strains of *Drosophila melanogaster*, leading to resistance.” | [Wiley article](https://onlinelibrary.wiley.com/doi/10.1002/ps.1472) | +| Supplementary source inventory, “Original [27]” | Optional completeness fix: add the retrospective publisher DOI `10.1016/S0022-3565(24)39005-6` to the Dawson entry. Its PMID record omits the DOI, but the publisher explicitly assigns it to the 2000 article. | [Journal of Pharmacology and Experimental Therapeutics](https://jpet.aspetjournals.org/article/S0022-3565%2824%2939005-6/abstract); [PubMed PMID 11082440](https://pubmed.ncbi.nlm.nih.gov/11082440/) | +| Main §2.6 [18]; source §2.6 [75] | Howard's paper establishes decision-theoretic information value; it does **not** derive `max(0, S_j) × Var(Y)`. Call this an **author-defined heuristic priority score**, motivated by information-value thinking, rather than formal value of information. | [IEEE article](https://ieeexplore.ieee.org/document/4082064/) | +| Source related work [67] | The statement that FlyBrainLab perturbations are necessarily “a hand-set multiplier with nothing behind it” is stronger than Lazar *et al.* establishes. Use the main manuscript's narrower wording: FlyBrainLab describes executable connectome-based circuits but not the specific per-row pharmacological evidence gate evaluated here. | [eLife article](https://elifesciences.org/articles/62362) | +| Main [5] dataset note | MaleCNS v1.0, the Cell paper, and the CC-BY licence are confirmed. I did **not** locate a primary Janelia release note supporting the exact “released 8 June 2026” date. Omit that day in the citation unless a primary release record is supplied. | [MaleCNS download](https://male-cns.janelia.org/download/); [PubMed PMID 42691995](https://pubmed.ncbi.nlm.nih.gov/42691995/) | + +**Other non-DOI records.** Main/source RatCVS [1] exists, and the repository's version history dates v3.3.5 to 19 June 2019 ([RatCVS source](https://github.com/johndempster/RatCVS/blob/master/RATMAIN.PAS)). Source [9]'s [FlyConnectome neurotransmitter ground-truth repository](https://github.com/flyconnectome/drosophila_neurotransmitters) exists; GitHub records its creation on 11 April 2024. Main [9]/source [10] is the dated [Mineault newsletter essay](https://www.neuroai.science/p/are-flies-playing-beat-saber), which is correctly labeled non-peer-reviewed. Source [45] [Greco *et al.*, PMID 7568331](https://pubmed.ncbi.nlm.nih.gov/7568331/) and [48] [Loewe, PMID 13081480](https://pubmed.ncbi.nlm.nih.gov/13081480/) match their cited author/title/journal/year/pages. + +**Claim-support and verification limits.** Shiu *et al.* [main 8/source 7] supports a connectome-based LIF model of **central-brain** feeding/grooming circuits and says all but one model parameter came from previous modelling/electrophysiology or the connectome; it is not a whole-MaleCNS or pharmacology validation ([Nature article](https://www.nature.com/articles/s41586-024-07763-9)). Mindlin *et al.* [main 11/source 68] reports that maximum simulated perturbation distance correlated with the **mean density of normalized receptor maps**, while a local node-strength comparison did not show the same result. This does not prove location never matters in receptor-weighted network models ([Communications Biology article](https://www.nature.com/articles/s42003-024-06852-9)). The provenance and interpretation of each numeric library row remain a separate row-by-row scientific audit; checking bibliography metadata cannot validate an inferred transfer, a binding-to-effect conversion, or a concentration value. diff --git a/flylab/analysis/baselines.py b/flylab/analysis/baselines.py index 52681e8..28f5060 100644 --- a/flylab/analysis/baselines.py +++ b/flylab/analysis/baselines.py @@ -1125,14 +1125,14 @@ def _statements(gain: dict[str, Any], scope: str) -> list[str]: else "" ) out.append( - f"{lvl} REPRODUCES the full model {scope}{unit_note}: " + f"{lvl} meets the declared rank-agreement threshold {scope}{unit_note}: " f"rho = {rho:.3f}, r = {r:.3f} over {row['n_compounds']} " - "compounds. The layers above it add no ordering information " - "for this readout." + "compounds. This is agreement in the tested library, not independent " + "validation of either model." ) elif row["reproduces_full_ordering"]: out.append( - f"{lvl} reproduces the full model's ORDERING but not its values " + f"{lvl} meets the rank threshold but not the value-agreement threshold " f"{scope}: rho = {rho:.3f} (>= {REPRODUCES_RHO}) with r = " f"{r:.3f} (< {REPRODUCES_R}) over {row['n_compounds']} compounds." ) @@ -1144,8 +1144,8 @@ def _statements(gain: dict[str, Any], scope: str) -> list[str]: ) else: out.append( - f"{lvl} does NOT reproduce the full model {scope}: rho = " - f"{rho:.3f}, r = {r:.3f}; the layers above it are doing real work." + f"{lvl} is below the declared rank threshold {scope}: rho = " + f"{rho:.3f}, r = {r:.3f}." ) cmp_block = gain.get("generic_rule_comparison") if cmp_block and cmp_block.get("delta_rho") is not None: @@ -1401,23 +1401,16 @@ def reference(draw) -> tuple[list[float], dict[str, Any], float | None]: f"{sq['median']:.3f}, 5th-95th percentile [{sq['p05']:.3f}, " f"{sq['p95']:.3f}] over {sq['n']} draws. The observed " f"{observed:.3f} is at the {sparse_pct:.0f}th percentile - " - f"{verdict}. A rank correlation of this size between levels B and D " - "is therefore close to an algebraic identity: both are functions of " + f"{verdict}. A high correlation between levels B and D can arise " + "because both are functions of " "the same gain vector, one through a fixed census and one through a " "row-normalised matrix whose cells average their presynaptic gains." ) statements.append( - "WHAT WOULD HAVE BEEN EVIDENCE AGAINST THE CLAIM that composition " - "carries the ordering: an observed rho below the matched " - f"reference's 5th percentile ({sq['p05']:.3f}), i.e. the census " - "ordering the REAL library worse than it orders pseudo-compounds " - "with no pharmacology in them. Evidence that the agreement was a " - "substantive finding about this connectome would need the matched " - "reference to sit well below the observed value - a median near " - "zero, so that reproducing the ordering was a priori unlikely. " - f"Here the matched reference median is {sq['median']:.3f}, so the " - "quoted 0.99 should be read as the structural floor plus a small " - "excess, not as a measurement of what the connectome contributes." + "The observed value should be read against the matched reference " + f"interval [{sq['p05']:.3f}, {sq['p95']:.3f}], whose median is " + f"{sq['median']:.3f}. A large B-versus-D correlation alone does not " + "measure a compound-specific contribution of the connectome." ) if shuffled is not None and observed is not None: statements.append( @@ -1426,9 +1419,8 @@ def reference(draw) -> tuple[list[float], dict[str, Any], float | None]: f"(difference {shuffled - observed:+.3g}). Both levels are " "functions of the gain vector alone, so a permutation of the labels " "moves the (B, D) pairs as a block and cannot change the rank " - "correlation. The comparison therefore carries no compound-level " - "information: it is a statement about the map from gains to " - "readouts, not about which compound is which." + "correlation. This label shuffle supplies no independent " + "compound-level evidence; it tests the shared gain-to-readout map." ) return { "conc_M": float(conc_M), @@ -1548,7 +1540,7 @@ def composition_dominance_under_normalisations( else f"survives only under {survives or 'none'}" ) + f" (threshold rho >= {REPRODUCES_RHO}); the shipped engine is " - f"{ref!r}, and the row normalisation is an undocumented modelling " + f"{ref!r}, and the row normalisation is a documented modelling " "choice, not a measurement." ) return { diff --git a/flylab/analysis/claims.py b/flylab/analysis/claims.py index 9144a63..b41ef90 100644 --- a/flylab/analysis/claims.py +++ b/flylab/analysis/claims.py @@ -96,13 +96,17 @@ "id": "mn9_receptor_expression", "statement": "The receptor complement of MN9 -- the main motor readout -- is not measured.", "why": ( - "No cited atlas reports Rdl or nAChR expression for identified adult leg or " - "labellar motor neurons, so MN9 carries the 'unknown' default weight and the " - "model patches it with the same uniform gain as every other cell." + "No class-specific receptor fractions for identified adult leg or labellar motor " + "neurons have been added to this dataset. Adult VNC primary counts (GEO GSE141807) " + "have been reanalyzed for VNC-wide gene detection, but the summary has no " + "cell-class assignments and does not identify MN9 in the brain subesophageal zone. " + "MN9 therefore carries the 'unknown' default weight, and the model patches it with " + "the same uniform gain as every other cell." ), "would_resolve": ( - "Adult VNC snRNA-seq restricted to annotated motor-neuron types, reporting " - "nAChR / Rdl / GluCl / Ace expression." + "A defensible mapping from the reprocessed GEO GSE141807 barcodes to adult VNC " + "motor-neuron classes, plus a separate class-matched brain/SEZ atlas or targeted " + "assay for MN9." ), }, { @@ -430,13 +434,19 @@ def claim_audit(notebook_or_result: Mapping[str, Any] | None = None, **kw: Any) "MODEL-ASSUMPTION", "MODEL-ASSUMPTION", ( - "The gain is applied uniformly to every cell of a transmitter class, because " - "per-cell receptor expression is largely unmapped" - + (f" ({coverage * 100:.1f}% of cells mapped)" if coverage is not None else "") - + ". A cell that does not express the receptor is patched exactly like one that does." + "The gain is applied uniformly to every cell of a transmitter class. Coarse " + "cross-atlas class-level annotation is available for" + + ( + f" {coverage * 100:.1f}% of graph-node × receptor-key pairs" + if coverage is not None + else " an unquantified share of graph-node × receptor-key pairs" + ) + + "; this is annotation availability, not a per-cell expression measurement. " + "A cell that does not express the receptor is patched exactly like one that does." ), detail={ - "fraction_cells_mapped": coverage, + "fraction_cell_receptor_pairs_annotated": coverage, + "coverage_unit": "graph_node_x_receptor_key_pair", "motor_neuron_gap": mn9_gap, "primary_model": "uniform gains", }, @@ -818,7 +828,12 @@ def _expression_facts() -> tuple[float | None, dict[str, Any] | None]: table = expression_table() except Exception: # pragma: no cover - optional module return None, None - coverage = _num((table.get("coverage") or {}).get("fraction_known_overall")) + coverage_block = table.get("coverage") or {} + coverage = _num( + coverage_block.get("fraction_cell_receptor_pairs_annotated") + if "fraction_cell_receptor_pairs_annotated" in coverage_block + else coverage_block.get("fraction_known_overall") + ) gap = table.get("motor_neuron_gap") slim = None if isinstance(gap, Mapping): @@ -826,6 +841,7 @@ def _expression_facts() -> tuple[float | None, dict[str, Any] | None]: "status": gap.get("status"), "description": gap.get("description"), "consequence_for_flylab": gap.get("consequence_for_flylab"), + "public_data_lead": gap.get("public_data_lead"), } return coverage, slim diff --git a/flylab/analysis/dependence.py b/flylab/analysis/dependence.py index a722bfd..4debc8f 100644 --- a/flylab/analysis/dependence.py +++ b/flylab/analysis/dependence.py @@ -963,8 +963,10 @@ def necessary_information_level( This function used to report "the weakest graph model that already reproduces the effect". It no longer does, and no string it returns says "reproduces": failing to reject a null is not evidence of equivalence. - When every null is distinguishable the answer is ``real_connectome``: - nothing less than the MaleCNS cut is indistinguishable from it. + When every ranked information-ladder rung is distinguishable the answer is + ``real_connectome``: nothing less than the MaleCNS cut is indistinguishable + from it. The orthogonal plain ``sign_permute`` joint target-and-sign null is + not part of that ladder and does not determine this answer. """ if isinstance(profile, dict): rows = list(profile.get("modes") or []) @@ -1432,7 +1434,7 @@ def _assemble_profile( "confirmatory tests (" + ", ".join(CONFIRMATORY_COMPOUNDS) + f" at n >= {PAPER_N}) and carries no multiplicity correction of " "its own. Counting verdicts over many such profiles needs the " - "FDR-controlled dependence_landscape instead." + "component-test-adjusted dependence_landscape instead." ) profile = { "compound": compound, @@ -1977,7 +1979,7 @@ def dependence_landscape( n_jobs: int = 1, **kw: Any, ) -> dict[str, Any]: - """The compound x concentration dependence landscape, FDR-controlled. + """The compound x concentration dependence landscape with component-test BH adjustment. One cell per ``(compound, conc_M)``: its real effect, the per-mode permutation p and z, the three-way verdict, the dependence class and the @@ -2540,17 +2542,19 @@ def _landscape_result( topo_raw = sorted({r["compound"] for r in table if r["class_raw"] == "topology-dependent"}) comp = sorted({r["compound"] for r in table if r["class"] == "composition-dominated"}) statement = ( - "exploratory screen, FDR-controlled: " + "exploratory screen with BH-adjusted structural component tests: " f"{n_topo} of {len(cells)} cells are topology-dependent after " f"Benjamini-Hochberg at fdr_alpha = {q_alpha} across {bh['m']} structural " - f"tests ({n_topo_raw} before correction)." + f"tests ({n_topo_raw} before correction). The derived cell classes have " + "no separate cell-level FDR guarantee." if not confirmatory else ( - f"confirmatory-resolution landscape: {n_topo} of {len(cells)} cells " + f"higher-resolution landscape: {n_topo} of {len(cells)} cells " f"are topology-dependent after Benjamini-Hochberg at fdr_alpha = " f"{q_alpha} across {bh['m']} structural tests ({n_topo_raw} before " "correction); the permutation resolution supports the adjusted " - "threshold for a single isolated test." + "threshold for a single isolated test. The derived cell classes have " + "no separate cell-level FDR guarantee." ) ) if not bh["can_reject"]: diff --git a/flylab/analysis/nullmodels.py b/flylab/analysis/nullmodels.py index 0d3cee0..6240bd8 100644 --- a/flylab/analysis/nullmodels.py +++ b/flylab/analysis/nullmodels.py @@ -1369,8 +1369,9 @@ def connectome_information_score( ``panel`` may be a :func:`null_panel` result or its ``rows``; when it is ``None`` the panel is computed first. ``assay`` filters the rows to one - assay arm (default: all rows). 1.0 means the connectome mattered against - every null; 0.0 means a shuffled graph reproduced the drug effect. + assay arm (default: all rows). 1.0 means the real graph was distinguishable + from every included null at the selected threshold; 0.0 means none was. + This legacy summary does not establish equivalence when a null is not rejected. .. deprecated:: 0.6 diff --git a/flylab/analysis/voi.py b/flylab/analysis/voi.py index 675b6bc..f202a9e 100644 --- a/flylab/analysis/voi.py +++ b/flylab/analysis/voi.py @@ -195,18 +195,20 @@ "experiment": "FISH / scRNA-seq for receptor expression in the named cell class", "design": ( "HCR-FISH for nAChR subunits, Rdl and GluCl on identified adult " - "labellar motor neurons (MN9) and the labellar GRN types, or a " - "MaleCNS-registered single-cell atlas that carries superclass labels." + "labellar motor neurons (MN9) and the labellar GRN types; first " + "determine whether adult VNC motor-neuron labels can be joined to the " + "reprocessed GEO GSE141807 barcodes, then use a separate " + "MaleCNS-registered brain/SEZ atlas or targeted assay for MN9." ), "resolves": ( - "which cells actually carry the receptor the drug acts on; FlyLab's " - "expression coverage is 0.21 and adult motor-neuron receptor " - "expression -- MN9, the main readout -- has no source at all" + "which cells actually carry the receptor the drug acts on; only 0.205 of " + "graph-node × receptor-key pairs currently have a coarse class-level annotation, " + "and MN9-specific receptor expression remains unresolved" ), - "why_now": "a confirmed literature gap, and a data problem rather than a modelling one", + "why_now": "VNC-wide counts are reprocessed; class mapping and an MN9 brain/SEZ match remain", "cost": "medium", - "feasibility": "standard method; the MN9 result would be new", - "blocking_gate": "v0.6 gate 1 (expression coverage)", + "feasibility": "VNC barcode-to-class mapping needs assessment; an MN9-specific result may require new data", + "blocking_gate": "v0.7 gate 7 (receptor-expression evidence)", }, "weight_threshold": { "experiment": "no experiment: a reconstruction-confidence analysis", diff --git a/flylab/browser/bridge.py b/flylab/browser/bridge.py index 17d3411..fe35686 100644 --- a/flylab/browser/bridge.py +++ b/flylab/browser/bridge.py @@ -1760,14 +1760,21 @@ def _trust_block( { "area": "Receptor expression", "status": ( - f"{coverage * 100:.1f}% of cells mapped" if coverage is not None else "unmapped" + f"{coverage * 100:.1f}% of graph-node × receptor-key pairs annotated" + if coverage is not None + else "annotation coverage unavailable" ), "classification": "MODEL-ASSUMPTION", "basis": ( - "Gains are applied uniformly. Adult motor-neuron receptor expression -- the " - "class MN9 belongs to -- is a confirmed gap in the literature." + "The percentage is the share of graph-node × receptor-key pairs with a coarse " + "cross-atlas class annotation, not the share of cells measured to express a " + "receptor. Gains remain uniform. MN9-specific receptor expression is unresolved." ), - "detail": {"fraction_cells_mapped": coverage, "motor_neuron_gap": mn9_gap}, + "detail": { + "fraction_cell_receptor_pairs_annotated": coverage, + "coverage_unit": "graph_node_x_receptor_key_pair", + "motor_neuron_gap": mn9_gap, + }, }, { "area": "Live validation", @@ -1797,11 +1804,18 @@ def _expression_coverage() -> tuple[float | None, dict[str, Any] | None]: table = expression_table() except Exception: # pragma: no cover - optional module return None, None - coverage = (table.get("coverage") or {}).get("fraction_known_overall") + coverage_block = table.get("coverage") or {} + coverage = coverage_block.get("fraction_cell_receptor_pairs_annotated") + if coverage is None: + coverage = coverage_block.get("fraction_known_overall") gap = table.get("motor_neuron_gap") slim = None if isinstance(gap, dict): - slim = {"status": gap.get("status"), "description": gap.get("description")} + slim = { + "status": gap.get("status"), + "description": gap.get("description"), + "public_data_lead": gap.get("public_data_lead"), + } try: coverage = float(coverage) except (TypeError, ValueError): diff --git a/flylab/cli.py b/flylab/cli.py index 8e06b8a..4367696 100644 --- a/flylab/cli.py +++ b/flylab/cli.py @@ -1434,9 +1434,11 @@ def _tutorial_own_compound() -> None: "note": ( "Same compound, same dose, same readout, opposite verdict. The " "paper-scale command is the same line with --n 1000. The scale " - "ladder (flylab.analysis.scale.available_cuts) moves it again with " - "size alone: composition-dominated at ~1100 cells, distinguishable " - "from every null at ~5000. See docs/INTERPRETATION.md section 4." + "saved scale-ladder results are consistent with recurrence, not node " + "count, tracking the change; this does not isolate a causal feature. " + "At 5k the imidacloprid effect is distinguishable from the three " + "ranked edge/wiring nulls. The exact large cuts are not present in " + "every checkout. See docs/INTERPRETATION.md section 4." ), }, ], diff --git a/flylab/pharm/expression.py b/flylab/pharm/expression.py index f12ed8a..13b7248 100644 --- a/flylab/pharm/expression.py +++ b/flylab/pharm/expression.py @@ -243,35 +243,74 @@ def coverage( aggregation: str = "mean", path: Path | None = None, ) -> dict[str, Any]: - """How many cells of the loaded graphs sit in classes with known expression.""" + """Summarize class-annotation coverage, never cell-level expression. + + ``fraction_cell_receptor_pairs_annotated`` is the share of graph-node × + receptor-key pairs whose superclass has any known class-level statement + in the cross-atlas dataset. It does not estimate the fraction of cells + expressing a receptor. The legacy ``fraction_known_overall`` key remains + as an alias for compatibility. + """ from flylab.circuit.rate import GRAPHS names = list(graphs) if graphs is not None else list(GRAPHS) receptors = [receptor] if receptor else list(RECEPTOR_GENES) per_graph: dict[str, Any] = {} - totals = {"n_cells": 0, "n_known": 0} + totals = { + "n_graph_nodes": 0, + "n_nodes_with_any_annotation": 0, + "n_cell_receptor_pairs": 0, + "n_annotated_cell_receptor_pairs": 0, + } for name in names: try: classes = _graph_superclasses(name) except Exception: # pragma: no cover - graph not committed in this checkout continue block: dict[str, Any] = {"n_cells": len(classes)} + annotated_superclasses: set[str] = set() for rec in receptors: index = _superclass_index(rec, None, default_unknown, aggregation, path) known = sum(1 for sc in classes if sc and index.get(sc, {}).get("known")) + annotated_superclasses.update( + sc for sc in classes if sc and index.get(sc, {}).get("known") + ) block[rec] = { "n_known": known, "n_unknown": len(classes) - known, "fraction_known": known / len(classes) if classes else 0.0, } - totals["n_known"] += known - totals["n_cells"] += len(classes) + totals["n_annotated_cell_receptor_pairs"] += known + totals["n_cell_receptor_pairs"] += len(classes) + block["n_cells_with_any_class_annotation"] = sum( + 1 for sc in classes if sc and sc in annotated_superclasses + ) + block["fraction_cells_with_any_class_annotation"] = ( + block["n_cells_with_any_class_annotation"] / len(classes) if classes else 0.0 + ) + totals["n_nodes_with_any_annotation"] += block["n_cells_with_any_class_annotation"] + totals["n_graph_nodes"] += len(classes) per_graph[name] = block - overall = totals["n_known"] / totals["n_cells"] if totals["n_cells"] else 0.0 + pair_fraction = ( + totals["n_annotated_cell_receptor_pairs"] / totals["n_cell_receptor_pairs"] + if totals["n_cell_receptor_pairs"] + else 0.0 + ) return { "graphs": per_graph, "receptors": receptors, - "fraction_known_overall": overall, + "fraction_cell_receptor_pairs_annotated": pair_fraction, + "n_annotated_cell_receptor_pairs": totals["n_annotated_cell_receptor_pairs"], + "n_cell_receptor_pairs": totals["n_cell_receptor_pairs"], + "fraction_graph_nodes_with_any_class_annotation": ( + totals["n_nodes_with_any_annotation"] / totals["n_graph_nodes"] + if totals["n_graph_nodes"] + else 0.0 + ), + "n_graph_nodes": totals["n_graph_nodes"], + # Backwards-compatible alias; this has always been a cell-receptor-pair fraction. + "fraction_known_overall": pair_fraction, + "coverage_unit": "graph_node_x_receptor_key_pair", "aggregation": aggregation, "default_unknown": default_unknown, "warning": EXPRESSION_WARNING, diff --git a/flylab/pharm/library.yaml b/flylab/pharm/library.yaml index 259fe54..523a8ad 100644 --- a/flylab/pharm/library.yaml +++ b/flylab/pharm/library.yaml @@ -48,7 +48,7 @@ notes: measured_fit: Reserved. Only for values fitted to data imported by the user. Nothing in the shipped library uses this tier. hill_n: Hill coefficients are teaching defaults in the 1.0-1.5 range unless the source reports steeper cooperativity. efficacy: Optional relative efficacy (0..1), used as the tau-like term of the operational model in flylab.pharm.binding. Absent means 'assume full'. - nicotinic_subunits: 'insect_nAChR is a documented AGGREGATE key and still drives the g_ach gain rules. insect_nAChR_alpha6 and insect_nAChR_beta1 carry the subunit-resolved rows: spinosad''s value sits on alpha6 (Perry 2015) and imidacloprid''s aphid binding Kd sits on beta1 (Bass 2011, R81T; cross-species/E1, so it reports binding_engagement_proxy). A third nicotinic key, insect_nAChR_native_dmel, is preparation-resolved rather than subunit-resolved and holds the one Drosophila-measured binding constant in the library (Tomizawa 1996). Neonicotinoid electrophysiology is nearly all done on hybrid Dalpha/chicken-beta2 or native mixed preparations, which is why those rows stay on the aggregate key and say so in relation/source. No number was invented for the split.' + nicotinic_subunits: 'insect_nAChR is a documented AGGREGATE key and still drives the g_ach gain rules. insect_nAChR_alpha6 and insect_nAChR_beta1 carry the subunit-resolved rows: spinosad''s target is alpha6 (Perry 2015), but its potency is currently unsupported and imidacloprid''s aphid binding Kd sits on beta1 (Bass 2011, R81T; cross-species/E1, so it reports binding_engagement_proxy). A third nicotinic key, insect_nAChR_native_dmel, is preparation-resolved rather than subunit-resolved and holds the one Drosophila-measured binding constant in the library (Tomizawa 1996). Neonicotinoid electrophysiology is nearly all done on hybrid Dalpha/chicken-beta2 or native mixed preparations, which is why those rows stay on the aggregate key and say so in relation/source. The alpha6 target row is withheld until a matched potency and direction are sourced.' v04_continuity: imidacloprid, nitenpyram, nicotine, diazepam and fipronil keep the exact value_M / n / direction shipped in v0.4 and v0.5; schema v3 only adds the type of each value, its relation and its species. Tests depend on those numbers. receptors: insect_nAChR: @@ -160,7 +160,7 @@ compounds: efficacy: 1.0 relation: exact_compound_related_receptor species: Ctenocephalides felis alpha1/alpha2 + Gallus gallus beta2 hybrid - source: 'Dederer et al. 2011, Insect Biochem Mol Biol 41:51 (Drosophila/Ctenocephalides alpha1-beta1 nAChR; neonicotinoid agonist potency in the 10-100 nM order). Value kept unchanged from FlyLab v0.4. SOURCE DESCRIPTION CORRECTED in v0.6 (schema v3): Dederer et al. expressed cat-flea Cfalpha1/Cfalpha2 with the CHICKEN beta2 subunit, i.e. a hybrid receptor - not an ''alpha1-beta1'' receptor as the v0.5 source string said. The value is unchanged; only the description of what was measured is corrected. It is an agonist EC50 on a hybrid receptor, so this row is typed EC50 / exact_compound_related_receptor (evidence distance E2) and drives functional_engagement_proxy, never binding occupancy. The subunit-resolved beta1 row is insect_nAChR_beta1.' + source: 'Dederer H, Werr M, Ilg T (2011) Insect Biochem Mol Biol 41(1):51-61. DOI 10.1016/j.ibmb.2010.09.012; PMID 20933086. Imidacloprid EC50 was 0.02 uM on cat-flea Cfalpha1/chicken Gg-beta2 hybrid receptors. The same paper also studied Drosophila alpha subunits with chicken beta2, but the 0.02 uM value is not a Drosophila receptor measurement. This unchanged library value is therefore an E2 related-receptor functional-engagement proxy, not binding occupancy.' evidence_tier: literature_order vertebrate_nAChR_a4b2: param_type: EC50 @@ -368,8 +368,8 @@ compounds: direction: antagonist efficacy: 1.0 relation: exact_compound_exact_receptor_other_species - species: Homo sapiens recombinant GABA-A ([3H]EBOB displacement) - source: 'Ratra and Casida 2001, Toxicol Lett 122:215 ([3H]EBOB displacement, human recombinant GABA-A): vertebrate average IC50 for fipronil is 1103 nM = 1.1e-6 M. CORRECTED in v0.5: FlyLab v0.4 carried 1.0e-05 M while citing this same paper, so the library contradicted its own citation; the value is now the cited one. SUBUNIT CAVEAT: the human beta3 HOMOpentamer is far more sensitive, 0.5-2.4 nM (Ratra, Kamita and Casida 2001, Toxicol Appl Pharmacol 172:233), so the ''vertebrate-safe'' reading of this row applies to native heteromeric receptors only, not to beta3 homomers.' + species: native GABA-related chloride-channel preparations from multiple vertebrate species ([3H]EBOB displacement) + source: 'Hainzl, Cole and Casida 1998, Chem Res Toxicol 11:1529-1535 (DOI 10.1021/tx980157t; PMID 9860498): mean fipronil IC50 1103 nM across human, dog, mouse, chicken, quail and salmon preparations by [3H]EBOB displacement. This is a multi-species binding-site proxy, not a human recombinant functional GABA-A potency.' evidence_tier: literature_order insect_RDL: param_type: IC50 @@ -377,9 +377,9 @@ compounds: n: 1.2 direction: antagonist efficacy: 1.0 - relation: exact_compound_exact_receptor_exact_species - species: Drosophila melanogaster - source: 'Hosie et al. 1995, Br J Pharmacol 115:909 (fipronil block of dieldrin-sensitive Drosophila GABA receptors, tens-of-nM order). SPREAD NOTE: the literature spans ~1500-fold for this one row - 2.4-12 nM on house-fly head membranes ([3H]EBOB displacement; Cole et al. 1993, Pestic Biochem Physiol 46:47 and Ratra and Casida 2001, Toxicol Lett 122:215) versus 0.3-3 uM on the predominant in vivo splice/edit variant of Drosophila RDL in Xenopus oocytes (Lees et al. 2014, PLoS ONE 9:e97468). Library keeps the potent end (unchanged from FlyLab v0.4); it is a defensible midpoint-to-potent choice, not a measured constant.' + relation: exact_compound_related_receptor + species: insect GABA-related chloride-channel preparations; Drosophila RDL variant response differs by assay + source: 'The 30 nM model value is an assay-transfer proxy, not a measured constant for the predominant Drosophila RDLbd I/V variant. Hainzl et al. 1998 (DOI 10.1021/tx980157t) reports 3-12 nM insect [3H]EBOB displacement; Lees et al. 2014 (DOI 10.1371/journal.pone.0097468) reports about 0.63 uM functional block on RDLbd I/V. Model sensitivity to this range is unresolved.' evidence_tier: literature_order clothianidin: name: Clothianidin @@ -546,15 +546,14 @@ compounds: notes: 'Teaching case: the electrophysiological EC50 on isolated neurons is far weaker than the whole-insect potency.' receptors: insect_nAChR: - param_type: EC50 - value_M: 0.00015 + param_type: unknown + value_M: null n: 1.0 - direction: agonist - efficacy: 1.0 - relation: exact_compound_related_receptor - species: Periplaneta americana DUM neurons (alpha-bungarotoxin-insensitive nAChR) - source: Watson et al. 2011, Pestic Biochem Physiol 101:32 (sulfoxaflor fully activates alpha-bungarotoxin-insensitive nAChRs of cockroach DUM neurons, EC50 about 154 uM); Sparks et al. 2013, Pestic Biochem Physiol 107:1 (mode of action review). - evidence_tier: literature_order + direction: none + efficacy: 0.0 + relation: unsupported + source: 'Placeholder pending source verification: Watson et al. 2011, Insect Biochem Mol Biol 41:432-439 (DOI 10.1016/j.ibmb.2011.01.009; PMID 21296156) supports Dalpha2/chicken beta2 receptor activity but does not verify the former 154 uM value in the checked record. A table-verified potency is required.' + evidence_tier: class_placeholder vertebrate_nAChR_a4b2: param_type: EC50 value_M: 0.001 @@ -596,18 +595,17 @@ compounds: name: Spinosad (spinosyn A) class: spinosyn cas: 131929-60-7 - notes: 'Allosteric activator: acts at a site distinct from the acetylcholine site and still activates alpha-bungarotoxin-blocked receptors.' + notes: 'Dalpha6 target involvement is documented, but the former 5 uM agonist EC50 was not supported by the cited study; this compound is unmodelled pending an assay-matched potency and direction.' receptors: insect_nAChR: - param_type: EC50 - value_M: 5.0e-06 + param_type: unknown + value_M: null n: 1.0 - direction: positive_modulator - efficacy: 1.0 - relation: exact_compound_related_receptor - species: Drosophila melanogaster Dalpha6 / heterologous insect alpha6-like receptors - source: 'Perry et al. 2015, Insect Biochem Mol Biol 64:106 (Drosophila Dalpha6 is required for spinosad action in vivo); heterologous alpha6 nAChR studies report allosteric activation in the low-micromolar order (EC50 about 5 uM). AGGREGATE ROW (v0.6): the measurement behind this number is alpha6-specific and is recorded subunit-resolved in insect_nAChR_alpha6. The same value is repeated here on the documented aggregate key (no new number was invented) so the legacy g_ach gain rule still fires; read as an aggregate it overstates the receptor population the value applies to.' - evidence_tier: literature_order + direction: none + efficacy: 0.0 + relation: unsupported + source: 'Placeholder pending source verification: Perry et al. 2015 (DOI 10.1016/j.ibmb.2015.01.017; PMID 25747008) supports Dalpha6 target involvement but not a 5 uM receptor EC50. Martelli et al. 2022 (PMID 35191376) reports low-dose antagonism of Dalpha6-mediated cholinergic response. No quantitative receptor potency or direction is admitted pending an assay-matched source.' + evidence_tier: class_placeholder vertebrate_nAChR_a4b2: param_type: unknown value_M: null @@ -645,15 +643,14 @@ compounds: source: class-order placeholder evidence_tier: class_placeholder insect_nAChR_alpha6: - param_type: EC50 - value_M: 5.0e-06 + param_type: unknown + value_M: null n: 1.0 - direction: positive_modulator - efficacy: 1.0 - relation: exact_compound_related_receptor - species: Drosophila melanogaster Dalpha6 (in vivo requirement) / heterologous insect alpha6-like receptors (the exact preparation behind the 5 uM order is not pinned) - source: 'Perry et al. 2015, Insect Biochem Mol Biol 64:106 (Drosophila Dalpha6 is required for spinosad action in vivo; Dalpha6 knockouts are highly resistant); heterologous alpha6 nAChR studies report allosteric activation in the low-micromolar order (EC50 about 5 uM). SUBUNIT-RESOLVED ROW (v0.6): this is the assay the spinosad number actually came from. The identical value also appears on the aggregate insect_nAChR key so the legacy gain rules keep working; no new number was invented for the split.' - evidence_tier: literature_order + direction: none + efficacy: 0.0 + relation: unsupported + source: 'Placeholder pending source verification: Dalpha6 is a spinosad target (Perry et al. 2015, DOI 10.1016/j.ibmb.2015.01.017), but that study does not measure the previously entered 5 uM receptor EC50; Martelli et al. 2022 (PMID 35191376) reports low-dose antagonism. This row is intentionally not modelled until potency and effect direction are sourced.' + evidence_tier: class_placeholder acetylcholine: name: Acetylcholine class: endogenous transmitter @@ -788,15 +785,14 @@ compounds: source: Ratra and Casida 2001, Toxicol Lett 122:215 (picrotoxinin blocks mammalian GABA-A at about 1 uM, i.e. no useful insect selectivity). evidence_tier: literature_order vertebrate_GlyR: - param_type: IC50 - value_M: 1.0e-05 + param_type: unknown + value_M: null n: 1.0 - direction: antagonist - efficacy: 1.0 - relation: class_extrapolation - species: mammalian homomeric glycine receptor - source: Ratra and Casida 2001, Toxicol Lett 122:215 and standard glycine receptor pharmacology (picrotoxin blocks homomeric GlyR in the micromolar order). Order-of-magnitude value. - evidence_tier: literature_order + direction: none + efficacy: 0.0 + relation: unsupported + source: 'Placeholder pending source verification: The previously cited Ratra and Casida GABA-receptor paper does not assay GlyR. A direct glycine-receptor picrotoxin preparation and potency are required before this row can be quantified.' + evidence_tier: class_placeholder insect_nAChR: param_type: unknown value_M: null @@ -875,15 +871,14 @@ compounds: source: Kane et al. 2000, PNAS 97:13949 (drug-resistant Drosophila show GluCl is the avermectin target); invertebrate GluCl activation is reported in the 0.1-30 nM order. evidence_tier: literature_order insect_RDL: - param_type: EC50 - value_M: 1.0e-06 + param_type: unknown + value_M: null n: 1.0 - direction: positive_modulator - efficacy: 0.5 - relation: exact_compound_exact_receptor_exact_species - species: Drosophila melanogaster RDL (predominant in vivo variant) in Xenopus oocytes - source: Lees et al. 2014, PLoS ONE 9:e97468 (ivermectin activates and potentiates the predominant in vivo RDL variant at sub-micromolar to micromolar). - evidence_tier: literature_order + direction: none + efficacy: 0.0 + relation: unsupported + source: 'Placeholder pending source verification: Lees et al. 2014 (DOI 10.1371/journal.pone.0097468) reports ivermectin inhibition of GABA-evoked RDLbd I/V currents (pIC50 6.33), with no current from 10 uM ivermectin alone. The former positive-modulator EC50 contradicted that assay; an inhibitory model mapping has not yet been validated.' + evidence_tier: class_placeholder vertebrate_GlyR: param_type: EC50 value_M: 1.0e-06 @@ -920,15 +915,14 @@ compounds: notes: Irreversible AChE inhibitor. The library treats fractional AChE occupancy as fractional enzyme inhibition, which ignores time-dependence and ageing. receptors: insect_AChE: - param_type: IC50 - value_M: 1.0e-08 + param_type: unknown + value_M: null n: 1.0 - direction: inhibitor - efficacy: 1.0 - relation: exact_compound_exact_receptor_exact_species - species: Drosophila melanogaster AChE (Ace) - source: Fournier and Mutero 1994, Comp Biochem Physiol C 108:19 (Drosophila AChE is inhibited by oxon organophosphates in the low-nM order and is more sensitive than mammalian AChE). - evidence_tier: literature_order + direction: none + efficacy: 0.0 + relation: unsupported + source: 'Placeholder pending source verification: Fournier and Mutero 1994 is a resistance review, not a primary Drosophila chlorpyrifos-oxon 10 nM IC50 assay. The previous number is withheld pending a primary enzyme assay with exposure time.' + evidence_tier: class_placeholder vertebrate_AChE: param_type: IC50 value_M: 3.0e-07 @@ -1043,15 +1037,14 @@ compounds: source: Burton et al. 2011, Insect Biochem Mol Biol 41:723 (Drosophila para tail-current EC50 0.40 uM for permethrin). evidence_tier: literature_order vertebrate_Nav1_x: - param_type: EC50 - value_M: 4.0e-05 + param_type: unknown + value_M: null n: 1.0 - direction: positive_modulator - efficacy: 0.3 - relation: class_extrapolation - species: Rattus norvegicus brain type IIA sodium channels (fold-difference statement) - source: Smith and Soderlund 1998, Neurotoxicology 19:823 (permethrin is more than 100-fold less potent on rat brain type IIA sodium channels than on insect para). - evidence_tier: literature_order + direction: none + efficacy: 0.0 + relation: unsupported + source: 'Placeholder pending source verification: Smith and Soderlund 1998 (PMID 9863771) studied cypermethrin, not permethrin. It cannot support the former 40 uM permethrin parameter; a matched mammalian permethrin assay is required.' + evidence_tier: class_placeholder insect_nAChR: param_type: unknown value_M: null diff --git a/flylab/static/app.js b/flylab/static/app.js index 19761e8..25f942e 100644 --- a/flylab/static/app.js +++ b/flylab/static/app.js @@ -402,6 +402,233 @@ }; } + /* Keep the primary bench controls and the saved experiment design together. + Editing one setting must not silently discard the experiment specification. */ + function persistDesign(extra) { + const saved = store("design") || {}; + store("design", Object.assign({}, saved, design(), extra || {})); + } + + const PRESET_FORMAT = "flylab.design-preset"; + const PRESET_VERSION = 1; + + function optionHas(select, value) { + return Array.from(select.options).some((option) => option.value === value); + } + + function selectedOptions(id, fallback) { + const values = $$("#" + id + " option:checked").map((option) => option.value); + return values.length ? values : fallback.slice(); + } + + function experimentSettings() { + const d = design(); + return { + assay: $("#f-exp-assay").value, + compounds: selectedOptions("f-exp-compounds", [d.compound].filter(Boolean)), + ladder_start_M: Number($("#f-exp-start").value), + ladder_stop_M: Number($("#f-exp-stop").value), + points_per_decade: Number($("#f-exp-ppd").value), + replicates: Number($("#f-exp-reps").value), + readouts: selectedOptions("f-exp-readouts", ["mn9_hz", "mean_hz"]), + randomize: $("#f-exp-randomize").checked, + blind: $("#f-exp-blind").checked, + include_vehicle: $("#f-exp-vehicle").checked, + }; + } + + function validateExperimentSettings(value) { + if (!value || typeof value !== "object" || Array.isArray(value)) throw new Error("experiment settings must be an object"); + const assaySelect = $("#f-exp-assay"); + const compoundsSelect = $("#f-exp-compounds"); + const readoutsSelect = $("#f-exp-readouts"); + if (typeof value.assay !== "string" || !optionHas(assaySelect, value.assay)) { + throw new Error("experiment assay is not available in this FlyLab build"); + } + const validList = (items, select, label) => { + if (!Array.isArray(items) || items.length === 0 || items.some((item) => typeof item !== "string" || !optionHas(select, item))) { + throw new Error("experiment " + label + " contains an unavailable value"); + } + if (new Set(items).size !== items.length) throw new Error("experiment " + label + " must not contain duplicates"); + return items.slice(); + }; + const compounds = validList(value.compounds, compoundsSelect, "compounds"); + const readouts = validList(value.readouts, readoutsSelect, "readouts"); + const positive = (v) => typeof v === "number" && Number.isFinite(v) && v > 0; + if (!positive(value.ladder_start_M) || !positive(value.ladder_stop_M)) { + throw new Error("concentration ladder start and stop must be positive finite numbers"); + } + if (!Number.isInteger(value.points_per_decade) || value.points_per_decade < 1 || value.points_per_decade > 10) { + throw new Error("points per decade must be an integer from 1 to 10"); + } + if (!Number.isInteger(value.replicates) || value.replicates < 1 || value.replicates > 64) { + throw new Error("replicates must be an integer from 1 to 64"); + } + ["randomize", "blind", "include_vehicle"].forEach((key) => { + if (typeof value[key] !== "boolean") throw new Error("experiment " + key + " must be true or false"); + }); + const decades = Math.abs(Math.log10(value.ladder_stop_M) - Math.log10(value.ladder_start_M)); + const concentrationCount = Math.max(1, Math.round(decades * value.points_per_decade)) + 1; + if (compounds.length * concentrationCount * value.replicates > 2000) { + throw new Error("experiment preset exceeds the 2,000-row design limit"); + } + return { + assay: value.assay, + compounds: compounds, + ladder_start_M: value.ladder_start_M, + ladder_stop_M: value.ladder_stop_M, + points_per_decade: value.points_per_decade, + replicates: value.replicates, + readouts: readouts, + randomize: value.randomize, + blind: value.blind, + include_vehicle: value.include_vehicle, + }; + } + + function validatePreset(value) { + if (!value || typeof value !== "object" || Array.isArray(value)) throw new Error("preset file must contain a JSON object"); + if (value.format !== PRESET_FORMAT) throw new Error("file is not a FlyLab design preset"); + if (value.version !== PRESET_VERSION) throw new Error("unsupported preset version; expected version " + PRESET_VERSION); + const raw = value.design; + if (!raw || typeof raw !== "object" || Array.isArray(raw)) throw new Error("preset is missing its main design settings"); + if (typeof raw.compound !== "string" || !optionHas($("#f-compound"), raw.compound)) { + throw new Error("preset compound is not available in this FlyLab build"); + } + if (typeof raw.conc_M !== "number" || !Number.isFinite(raw.conc_M) || raw.conc_M <= 0 || raw.conc_M > 1) { + throw new Error("preset concentration must be greater than 0 and at most 1 M"); + } + if (!["named", "taste_motor"].includes(raw.graph) || !optionHas($("#f-graph"), raw.graph)) { + throw new Error("preset graph is not available in this FlyLab build"); + } + if (!["rate", "lif"].includes(raw.engine) || !optionHas($("#f-engine"), raw.engine)) { + throw new Error("preset engine is not available in this FlyLab build"); + } + ["sugar_hz", "bitter_hz"].forEach((key) => { + if (typeof raw[key] !== "number" || !Number.isFinite(raw[key]) || raw[key] < 0 || raw[key] > 1000) { + throw new Error(key.replace("_hz", " drive") + " must be between 0 and 1,000 Hz"); + } + }); + if (!Number.isSafeInteger(raw.seed) || raw.seed < 0 || raw.seed > 2147483647) { + throw new Error("RNG seed must be an integer from 0 to 2,147,483,647"); + } + if (raw.genotype !== null && typeof raw.genotype !== "string") throw new Error("genotype must be a string or null"); + if (raw.genotype !== null && (!$("#genotype-field") || $("#genotype-field").hidden || !optionHas($("#f-genotype"), raw.genotype))) { + throw new Error("preset genotype is not available in this FlyLab build"); + } + return { + design: { + compound: raw.compound, + conc_M: raw.conc_M, + sugar_hz: raw.sugar_hz, + bitter_hz: raw.bitter_hz, + engine: raw.engine, + graph: raw.graph, + seed: raw.seed, + genotype: raw.genotype, + }, + experiment: validateExperimentSettings(value.experiment), + }; + } + + function applyExperimentSettings(settings) { + $("#f-exp-assay").value = settings.assay; + $("#f-exp-start").value = String(settings.ladder_start_M); + $("#f-exp-stop").value = String(settings.ladder_stop_M); + $("#f-exp-ppd").value = String(settings.points_per_decade); + $("#f-exp-reps").value = String(settings.replicates); + $$("#f-exp-compounds option").forEach((option) => { + option.selected = settings.compounds.includes(option.value); + }); + $$("#f-exp-readouts option").forEach((option) => { + option.selected = settings.readouts.includes(option.value); + }); + $("#f-exp-randomize").checked = settings.randomize; + $("#f-exp-blind").checked = settings.blind; + $("#f-exp-vehicle").checked = settings.include_vehicle; + updateLadderHint(); + } + + function presetForExport() { + const preset = { format: PRESET_FORMAT, version: PRESET_VERSION, design: design(), experiment: experimentSettings() }; + validatePreset(preset); + return preset; + } + + function presetStatus(message) { + const node = $("#preset-status"); + if (node) node.textContent = message; + } + + function applyPreset(preset) { + const validated = validatePreset(preset); + const main = validated.design; + $("#f-compound").value = main.compound; + setConc(main.conc_M); + $("#f-sugar").value = String(main.sugar_hz); + $("#f-bitter").value = String(main.bitter_hz); + $("#f-engine").value = main.engine; + $("#f-graph").value = main.graph; + $("#f-seed").value = String(main.seed); + if (!$("#genotype-field").hidden) $("#f-genotype").value = main.genotype || ""; + updateCompoundBadges(); + updateGraphHint(); + const minw = $("#f-cy-minw"); + if (minw) minw.value = main.graph === "taste_motor" ? 12 : 5; + applyExperimentSettings(validated.experiment); + + state.experiment = null; + state.runOrder = null; + state.blindKey = null; + state.revealed = false; + $("#btn-exp-csv").disabled = true; + $("#btn-exp-json").disabled = true; + $("#btn-exp-reveal").disabled = true; + $("#btn-exp-reveal").textContent = "Reveal codes"; + invalidate(); + renderExperiment(); + persistDesign({ experiment: experimentDesign(), experimentSettings: experimentSettings() }); + } + + function restoreExperimentSettings(saved) { + if (saved.experimentSettings) { + try { + applyExperimentSettings(validateExperimentSettings(saved.experimentSettings)); + return; + } catch (err) { + /* Invalid or stale browser storage falls back to the older saved design below. */ + } + } + const previous = saved.experiment; + if (previous) { + if (typeof previous.assay === "string" && optionHas($("#f-exp-assay"), previous.assay)) { + $("#f-exp-assay").value = previous.assay; + } + if (Number.isInteger(previous.replicates) && previous.replicates >= 1 && previous.replicates <= 64) { + $("#f-exp-reps").value = String(previous.replicates); + } + $$("#f-exp-compounds option").forEach((option) => { + option.selected = Array.isArray(previous.compounds) && previous.compounds.includes(option.value); + }); + const concentrations = Array.isArray(previous.concs_M) ? previous.concs_M.filter((value) => Number.isFinite(value) && value > 0) : []; + if (concentrations.length > 1) { + $("#f-exp-start").value = String(Math.min.apply(null, concentrations)); + $("#f-exp-stop").value = String(Math.max.apply(null, concentrations)); + } + if (Array.isArray(previous.readouts)) { + $$("#f-exp-readouts option").forEach((option) => { + option.selected = previous.readouts.includes(option.value); + }); + } + if (typeof previous.include_vehicle === "boolean") $("#f-exp-vehicle").checked = previous.include_vehicle; + updateLadderHint(); + } else { + $$("#f-exp-compounds option").forEach((option) => { + option.selected = option.value === $("#f-compound").value || option.value === "nicotine"; + }); + } + } + function setNotebook(nb, label) { if (!nb || typeof nb !== "object") return; state.notebook = nb; @@ -536,9 +763,18 @@ const saved = store("design") || {}; if (saved.compound && meta.compounds.some((c) => c.key === saved.compound)) sel.value = saved.compound; else if (meta.compounds.some((c) => c.key === "imidacloprid")) sel.value = "imidacloprid"; - if (saved.graph) $("#f-graph").value = saved.graph; - if (saved.engine) $("#f-engine").value = saved.engine; - if (saved.conc_M) setConc(Number(saved.conc_M)); + if (["named", "taste_motor"].includes(saved.graph)) $("#f-graph").value = saved.graph; + if (["rate", "lif"].includes(saved.engine)) $("#f-engine").value = saved.engine; + if (Number.isFinite(Number(saved.sugar_hz)) && Number(saved.sugar_hz) >= 0) { + $("#f-sugar").value = Math.min(1000, Number(saved.sugar_hz)); + } + if (Number.isFinite(Number(saved.bitter_hz)) && Number(saved.bitter_hz) >= 0) { + $("#f-bitter").value = Math.min(1000, Number(saved.bitter_hz)); + } + if (Number.isSafeInteger(Number(saved.seed)) && Number(saved.seed) >= 0) { + $("#f-seed").value = Number(saved.seed); + } + if (Number.isFinite(Number(saved.conc_M)) && Number(saved.conc_M) > 0) setConc(Number(saved.conc_M)); // re-apply the current concentration so the dashboard rail and its ladder // buttons show the right value on the first paint, saved design or not @@ -566,6 +802,9 @@ }) .join(""); $("#genotype-field").hidden = false; + if (saved.genotype && Array.from($("#f-genotype").options).some((o) => o.value === saved.genotype)) { + $("#f-genotype").value = saved.genotype; + } } } catch (err) { /* genotype module has not landed: the control stays hidden, by design */ @@ -1774,7 +2013,7 @@ const spec = experimentDesign(); const rows = spec.compounds.length * spec.concs_M.length * spec.replicates; if (rows > 2000) throw new Error(`design would make ${rows} rows; the server limit is 2000`); - store("design", Object.assign({}, design(), { experiment: spec })); + persistDesign({ experiment: spec, experimentSettings: experimentSettings() }); const out = await post("/api/experiment", spec); state.experiment = out; state.revealed = false; @@ -1811,7 +2050,17 @@ function renderExperiment() { const out = state.experiment; - if (!out) return; + if (!out) { + table( + "tbl-experiment", + [{ label: "compound" }, { label: "conc (M)", num: true }, { label: "n", num: true }], + [], + { empty: "Run the design to fill this grid." } + ); + draw("plot-exp-heat", [], { height: 220 }); + $("#exp-status").textContent = "Preset loaded; run the design to calculate these settings."; + return; + } const c = colors(); const readouts = (out.design && out.design.readouts) || ["mn9_hz"]; const primary = readouts[0]; @@ -3653,8 +3902,8 @@ notebook: { run: async () => { renderNotebook(); renderHistory(); return null; }, render: renderNotebook }, }; - /* Tabs that cost minutes of circuit time never start on their own. */ - const MANUAL_TABS = ["controls"]; + /* Expensive analyses and user-authored experiment designs never start on tab selection. */ + const MANUAL_TABS = ["controls", "experiment"]; let running = false; async function runTab(name) { @@ -3686,6 +3935,8 @@ function activate(name) { state.activeTab = name; + const mobilePicker = $("#mobile-panel-picker"); + if (mobilePicker) mobilePicker.value = name; $$('[role="tab"]').forEach((b) => b.setAttribute("aria-selected", String(b.id === "tab-" + name))); $$(".panel").forEach((p) => p.classList.toggle("active", p.id === "panel-" + name)); const panel = document.getElementById("panel-" + name); @@ -3742,36 +3993,73 @@ $$('[role="tab"]').forEach((b) => b.addEventListener("click", () => activate(b.id.replace("tab-", ""))) ); + const mobilePicker = $("#mobile-panel-picker"); + if (mobilePicker) mobilePicker.addEventListener("change", () => activate(mobilePicker.value)); $$("button[data-run]").forEach((b) => b.addEventListener("click", () => runTab(b.getAttribute("data-run"))) ); $("#btn-run").addEventListener("click", () => runTab(state.activeTab)); + $("#btn-preset-export").addEventListener("click", () => { + try { + const preset = presetForExport(); + download("flylab_design_preset_v1.json", JSON.stringify(preset, null, 2), "application/json;charset=utf-8"); + presetStatus("Exported FlyLab design preset version 1."); + } catch (err) { + presetStatus("Could not export: " + err.message); + toast("preset: " + err.message, "error"); + } + }); + $("#btn-preset-import").addEventListener("click", () => $("#f-preset-file").click()); + $("#f-preset-file").addEventListener("change", async (event) => { + const file = event.target.files && event.target.files[0]; + if (!file) return; + try { + if (file.size > 1024 * 1024) throw new Error("preset file is larger than 1 MB"); + const parsed = JSON.parse(await file.text()); + applyPreset(parsed); + presetStatus("Imported preset. Open Experiment and choose Run design to calculate results."); + toast("design preset imported", "ok"); + } catch (err) { + const message = (err && err.message) || "could not read preset file"; + presetStatus("Import failed: " + message); + toast("preset: " + message, "error"); + } finally { + event.target.value = ""; + } + }); + $("#f-logconc").addEventListener("input", (e) => { setConc(Math.pow(10, Number(e.target.value))); invalidate(); }); + $("#f-logconc").addEventListener("change", () => persistDesign()); $("#f-conc").addEventListener("change", (e) => { const v = Number(e.target.value); - if (Number.isFinite(v) && v > 0) setConc(v); - else toast("concentration must be a positive number in molar", "error"); + if (Number.isFinite(v) && v > 0) { + setConc(v); + persistDesign(); + } else { + toast("concentration must be a positive number in molar", "error"); + setConc(Math.pow(10, Number($("#f-logconc").value))); + } invalidate(); }); $("#f-compound").addEventListener("change", () => { updateCompoundBadges(); invalidate(); - store("design", design()); + persistDesign(); }); $("#f-graph").addEventListener("change", () => { updateGraphHint(); $("#f-cy-minw").value = $("#f-graph").value === "taste_motor" ? 12 : 5; invalidate(); - store("design", design()); + persistDesign(); }); ["#f-engine", "#f-seed", "#f-sugar", "#f-bitter", "#f-genotype"].forEach((sel) => $(sel).addEventListener("change", () => { invalidate(); - store("design", design()); + persistDesign(); }) ); @@ -3787,7 +4075,7 @@ rail.addEventListener("change", (e) => { setConc(Math.pow(10, Number(e.target.value))); invalidate(); - store("design", design()); + persistDesign(); runTab(state.activeTab); }); } @@ -3795,7 +4083,7 @@ b.addEventListener("click", () => { setConc(Number(b.getAttribute("data-conc"))); invalidate(); - store("design", design()); + persistDesign(); runTab(state.activeTab); }) ); @@ -3912,17 +4200,23 @@ } }); - ["#f-exp-start", "#f-exp-stop", "#f-exp-ppd"].forEach((sel) => - $(sel).addEventListener("input", updateLadderHint) - ); + ["#f-exp-start", "#f-exp-stop", "#f-exp-ppd"].forEach((sel) => { + $(sel).addEventListener("input", updateLadderHint); + $(sel).addEventListener("change", () => persistDesign({ experimentSettings: experimentSettings() })); + }); + ["#f-exp-compounds", "#f-exp-readouts", "#f-exp-assay", "#f-exp-reps", "#f-exp-vehicle"].forEach((sel) => { + $(sel).addEventListener("change", () => persistDesign({ experimentSettings: experimentSettings() })); + }); $("#f-exp-blind").addEventListener("change", () => { state.revealed = false; $("#btn-exp-reveal").disabled = !$("#f-exp-blind").checked || !state.experiment; renderExperiment(); + persistDesign({ experimentSettings: experimentSettings() }); }); $("#f-exp-randomize").addEventListener("change", () => { state.runOrder = null; renderExperiment(); + persistDesign({ experimentSettings: experimentSettings() }); }); $("#btn-exp-reveal").addEventListener("click", () => { state.revealed = !state.revealed; @@ -4317,19 +4611,8 @@ return; } - const saved = (store("design") || {}).experiment; - if (saved) { - if (saved.assay) $("#f-exp-assay").value = saved.assay; - if (saved.replicates) $("#f-exp-reps").value = saved.replicates; - $$("#f-exp-compounds option").forEach((o) => { - o.selected = (saved.compounds || []).indexOf(o.value) >= 0; - }); - updateLadderHint(); - } else { - $$("#f-exp-compounds option").forEach((o) => { - o.selected = o.value === $("#f-compound").value || o.value === "nicotine"; - }); - } + const saved = store("design") || {}; + restoreExperimentSettings(saved); $("#f-exp-reps").value = $("#f-exp-reps").value || 2; estimateCompare().catch(() => {}); diff --git a/flylab/static/index.html b/flylab/static/index.html index bd0e36f..cf2d4bb 100644 --- a/flylab/static/index.html +++ b/flylab/static/index.html @@ -82,6 +82,15 @@

Dose & design

Press R anywhere to run the open tab.
+
Main run settings are saved in this browser.
+
+
+ + +
+ +
Share these settings as a versioned JSON file.
+
@@ -107,6 +116,31 @@

Dose & design

+
+ + +
diff --git a/flylab/static/styles.css b/flylab/static/styles.css index d258af1..6b1bd35 100644 --- a/flylab/static/styles.css +++ b/flylab/static/styles.css @@ -323,6 +323,17 @@ kbd { color: var(--accent-ink); font-weight: 600; } +.mobile-panel-picker { display: none; } +.preset-tools { + display: grid; + gap: 5px; + margin-top: 12px; + padding-top: 10px; + border-top: 1px solid var(--border); +} +.preset-tools .row { gap: 6px; } +.preset-tools button { flex: 1; } +.preset-tools .hint { margin: 0; } .panel { display: none; } .panel.active { display: block; } @@ -564,8 +575,24 @@ pre.json { body { font-size: 13.5px; } .shell { padding: 10px; padding-bottom: 70px; } .grid, .grid.two { grid-template-columns: minmax(0, 1fr); } - .tabs { overflow-x: auto; flex-wrap: nowrap; } - [role="tab"] { white-space: nowrap; } + #main > .tabs { display: none; } + .mobile-panel-picker { + display: grid; + grid-template-columns: auto minmax(0, 1fr); + align-items: center; + gap: 10px; + margin-bottom: var(--gap); + padding: 9px 11px; + background: var(--surface); + border: 1px solid var(--border); + border-radius: var(--radius); + } + .mobile-panel-picker label { + color: var(--text-2); + font-size: 12px; + font-weight: 600; + } + .mobile-panel-picker select { width: 100%; min-width: 0; } #cy { height: 380px; } .topbar { padding: 8px 10px; } .tagline { display: none; } diff --git a/papers/IJRC_FlyLab_draft.md b/papers/IJRC_FlyLab_draft.md index d474b51..6f394e7 100644 --- a/papers/IJRC_FlyLab_draft.md +++ b/papers/IJRC_FlyLab_draft.md @@ -1,56 +1,55 @@ # Which Inputs Does a Network Simulation Depend On? Typed Evidence Propagation and Input-Dependence Testing, Evaluated on Cuts of a Synapse-Resolution *Drosophila* Connectome -**Research article, submitted draft for the International Journal of Research in Computing (IJRC)** +**Research article draft for the International Journal of Research in Computing (IJRC)** Software: https://github.com/pinkysworld/FlyLab -Version 0.6.0, commit `f95fd88a2165`, compound library SHA-256 `594b8d32c5e6`, map `male-cns:v1.0`. +Version 0.6.0, commit `50f5465b2df0`, compound library SHA-256 `359096467b33`, map `male-cns:v1.0`. > Every numeric span in this document is generated by `scripts/reproduce_paper.py` from `papers/results.json`; the prose is written around the numbers, not the other way round. No number in this paper was measured in a living animal. ## Abstract -A simulation built on a measured network and parameterised from published literature cannot, on its own, say which of its inputs its predictions depend on. We present a method that makes that question routine and testable, and an open implementation, FlyLab, that answers it per prediction. The method has four parts: **typed evidence propagation**, in which the parameter a source measured and its graded distance from the modelled target jointly decide which transformation is admissible, enforced by a soundness invariant checked exhaustively over the parameter library (56 of 102 rows refused rather than approximated); **input-dependence testing** against an ordered ladder of degraded copies of the network, reporting — with a prespecified equivalence margin and multiplicity control — the weakest input the test cannot distinguish from the real one; an **ablation ladder** over pipeline layers scored against a matched reference distribution; and **specification-family robustness** with variance attribution that ranks what to measure next. We evaluate the method rather than the domain: the dependence instrument is validated against planted ground truth with a power surface, its permutation budget is characterised by a convergence sweep, and the analysis is deterministic across two backends. **The principal result is methodological: a dependence verdict obtained on one extract of a network can invert on another extract of the same network, and the extract's recurrence, not its size, predicts which way.** Demonstrated on cuts of the public MaleCNS v1.0 *Drosophila* connectome, a prediction is composition-dominated on a sparse 1-hop extract (mean degree 1.208, 85% of edges onto four seed cells) and topology-dependent on every extract with real recurrence, including one with *fewer* nodes; across nested extracts of 1k, 5k, 10k, 25k, 50k cells the verdict settles as topology-dependent. Such a verdict is therefore a joint property of the prediction and the extract, and must be reported with the extract's structure. The central limitation is unchanged: the pipeline evaluates a model's behaviour, and its parameter-to-gain rules are asserted rather than measured. +Literature-parameterised network simulations can report predictions without showing which inputs carry them. We evaluate a method combining typed evidence gates, graph-degradation nulls, a matched ablation reference, gain-rule specifications and variance attribution on two cuts of the MaleCNS v1.0 *Drosophila* connectome. The current 21-compound library has 102 parameter rows; 63 do not yield numerical engagement after unsupported sources are withheld. This is a syntactic admission rule, not proof that every admitted assay value is correct: source-to-parameter verification remains incomplete. The incoming-star-like `named` cut has 85% of edges entering four seed cells. Under the declared floors and adjusted component tests, its 84 compound–concentration cells have 0 topology-dependent and 44 composition-dominated descriptive labels, while the denser `taste_motor` cut has 44 and 0. This contrast illustrates extract sensitivity, not a causal effect of recurrence. The gain transformation controls the direction of one model-output predicate. The larger scale study is imported from a saved record, and no independent circuit-level biological validation is available. -**Index terms:** computational method, scientific software, provenance, evidence typing, network simulation, null models, equivalence testing, multiple comparisons, specification robustness, sensitivity analysis, reproducible research, connectomics. +**Index terms:** computational method, scientific software, provenance, evidence typing, network simulation, null models, permutation testing, multiple comparisons, specification sensitivity, reproducible research, connectomics. ## 1. Introduction -A large class of scientific software has the same shape: a measured network, a table of parameters compiled from published literature, a rule that turns those parameters into perturbations of the network, and a numerical runtime that produces a prediction. Power grids, metabolic models, epidemic contact networks and connectome simulations are all built this way. Such a pipeline readily produces a number; what it does not produce is an answer to the question a reviewer asks first — *which of its inputs does this number actually depend on?* Without that answer the prediction cannot be audited, because a result carried entirely by an asserted coefficient is indistinguishable, at the output, from one carried by the measured structure. +A large class of scientific software has the same shape: a measured network, a table of parameters compiled from literature and declared model assumptions, a rule that turns those parameters into perturbations of the network, and a numerical runtime that produces a prediction. Related pipelines occur in several scientific domains; this study tests one connectome application. Such a pipeline readily produces a number; what it does not produce is an answer to the question a reviewer asks first — *which of its inputs does this number actually depend on?* Without that answer the prediction cannot be audited, because a result carried entirely by an asserted coefficient is indistinguishable, at the output, from one carried by the measured structure. Three properties make the question hard to answer in practice. The parameters are heterogeneous: sources measure different quantities, in different preparations, and are often simply absent, so a pipeline that coerces them into one numeric field silently manufactures evidence. The network enters through a runtime whose own conventions — normalisation, sign, iteration — are modelling choices that can carry part of the answer. And the usual test of whether structure matters is a permutation against a shuffled network, whose *non-rejection* is routinely written up as though the shuffle had reproduced the effect. -This paper contributes a method addressing all three (§2), an implementation that runs it per prediction rather than once per paper, and — the part we think is missing from this literature — an evaluation of the instruments themselves (§3). We demonstrate the method on cuts of a public synapse-resolution *Drosophila* connectome, a domain supplying exactly the hard case: a measured network, a literature parameter table with large gaps, and asserted transformation rules. The headline finding is about the method rather than the domain: **a dependence verdict obtained on one extract of a network can invert on another extract of the same network.** +This paper contributes a method addressing all three (§2), an implementation that runs it per prediction rather than once per paper, and — the part we think is missing from this literature — an evaluation of the instruments themselves (§3). We demonstrate the method on cuts of a public synapse-resolution *Drosophila* connectome, a domain supplying exactly the hard case: a measured network, a literature parameter table with large gaps, and asserted transformation rules. The central contribution is an audit of evidence admission, graph-null preservation and specification propagation. The two cuts illustrate how a descriptive dependence label changes with extraction. -We organise the work around four research questions. +We organise the scientific evaluation around three research questions; reproducibility is engineering support. * **RQ1.** Can heterogeneous published parameters be turned into computational perturbations *without* erasing their provenance or their parameter semantics, and can that discipline be stated as a checkable invariant rather than a convention? * **RQ2.** For a given prediction, which inputs does it depend on — and is that answer a property of the prediction, or of the extract of the network it was measured on? * **RQ3.** Which conclusions survive alternative admissible specifications, and which assumptions dominate the output's variance? -* **RQ4.** Can the entire analysis be reproduced identically, across backends, from auditable artifacts? -The answers are: yes, by grading the evidence, refusing invalid transformations and checking the refusal exhaustively; **it is a property of the extract**, which is the paper's main result; not all of them, and an undocumented normalisation inside the runtime carries part of RQ2's answer; and yes. An earlier version of this work answered RQ2 with "most predictions do not need the network", measured on a single extract — a conclusion this revision withdraws and explains. +The results show a finite, checkable admission rule for RQ1; extract- and runtime-dependent labels for RQ2; and specification-dependent model-output predicates for RQ3. These findings are conditional on the current parameter table and selected cuts. The RQ1 tests check declared types, readout paths and gain vectors; they do not verify the primary assay value behind each admitted row. ### 1.1 Related work and what is new here -**Network runtimes without parameterised perturbation.** Teaching simulators [1]–[4] model dose–response with no network at all. At the other extreme, FlyWire [5] and MaleCNS v1.0 [6] publish netlists with per-node annotations [10], [12]; Shiu *et al.* [7] show a literature-parameterised model over such a netlist reproduces a sensorimotor control, and downstream ports repeat the exercise with varying rigour [11]. FlyBrainLab [68] builds executable circuits directly from connectome data, but a perturbation, where present, enters as a hand-set multiplier with nothing behind it — exactly the case typed propagation is meant to replace. +**Network runtimes without parameterised perturbation.** RatCVS [1] and Virtual Rat Web [2] lack connectomes; PharmVR [3] is a preprint; [4] evaluates teaching. At the other extreme, FlyWire [5] and MaleCNS v1.0 [6] publish netlists with per-node annotations [9], [11]; Shiu *et al.* [7] show a literature-parameterised model over such a netlist reproduces a sensorimotor control, and downstream ports repeat the exercise with varying rigour [10]. FlyBrainLab [67] builds executable circuits from connectome data but does not describe the per-row pharmacological admission rule evaluated here. -**Parameter-map-informed network models.** On human structural connectomes, intervention is simulated by scaling regional responses with PET density maps [69], [70] — the closest published relative, whose most relevant result is also a negative one: in Mindlin *et al.* [69] the simulated improvement tracked the *mean* density of the activated receptors rather than where they sit. Methodologically, §2.4 is a multiverse analysis [74], [75] applied to a simulation, and §2.6 uses Saltelli cross-sampling [71] with Jansen estimators [72] validated on Ishigami [73] and ranked by value of information [76]. +**Parameter-map-informed network models.** On human structural connectomes, intervention is simulated by scaling regional responses with PET density maps [68], [69]. In Mindlin *et al.* [68], maximum simulated perturbation distance correlated with mean normalised receptor density, whereas their tested receptor-map/node-strength divergence metric did not. Methodologically, §2.4 is a multiverse analysis [73], [74] applied to a simulation, and §2.6 uses Saltelli cross-sampling [70] with Jansen estimators [71] validated on Ishigami [72] and ranked by an author-defined variance-priority heuristic motivated by information-value theory [75]. -**The domain evidence base.** The parameter library, and the supporting datasets for resistance alleles, expression, mixtures, exposure and behavioural endpoints, are compiled from the insect-versus-vertebrate pharmacology literature [8], [9], [13]–[67] and documented in §S4 and §S6; what matters methodologically is that it is typical of such tables — incomplete, heterogeneous in what each source measured, and internally inconsistent with its own citations in ten places that `data/literature/README.md` records without editing the library to match. Those gaps are the reason the type system changes results rather than documentation. +**The domain evidence base.** The parameter library combines published measurements, transferred values and explicit model assumptions [8], [12]–[66]. A row-level source audit found assay mismatches and unsupported numeric values; corrected rows are withheld from numerical engagement, while other exact values remain under primary-source review. The type system enforces declared categories but cannot itself establish that a paper supports the entered potency. -**Position.** Table I states what is new narrowly. The transferable part is not the domain coupling but the instrumentation around it: routine, per-prediction, *validated* testing of which inputs a prediction depends on, with non-rejection separated from equivalence, multiplicity controlled, and the test's own power characterised against planted ground truth. +**Position.** Table I states what is new narrowly. The transferable part is not the domain coupling but the instrumentation around it: routine per-prediction diagnostics with non-rejection distinguished from a descriptive point-gap pass, adjusted component tests and limited positive-control experiments. These are not a cell-level false-discovery guarantee or a general power estimate. -**Table I. What each system provides.** +**Table I. Capabilities described in the cited implementations.** A negative entry means the cited work does not report the capability; it is not a claim about every implementation in the category. -| Capability | Teaching sims [1]–[4] | Connectome LIF runtimes [7], [11] | FlyBrainLab [68] | Receptor-map whole-brain models [69], [70] | FlyLab | +| Capability | Teaching simulators [1]–[4] | Connectome LIF model [7] | FlyBrainLab [67] | Regional receptor models [68], [69] | This study | |---|---|---|---|---|---| | Named cells at synapse resolution | no | yes | yes | no (regions) | yes | | Compound **and free concentration** as the input | yes (organ level) | no | no | partly (receptor-map scaling) | yes | | Typed evidence with per-row provenance and distance | no | no | no | no | yes | | Two species scored at the same dose | no | no | no | no | yes | -| Dependence of each prediction on its inputs, tested | no | rarely | no | partly [69] | yes, routinely | -| The dependence test itself validated against ground truth | n/a | no | no | no | yes (§3.1, F16) | -| Specification robustness of the stated conclusions | no | no | no | no | yes | +| Dependence of each prediction on its inputs, tested | not reported | not reported in [7] | not reported | partly [68] | yes, for declared nulls | +| Synthetic recovery cases evaluated | n/a | not assessed | not assessed | not assessed | limited positive controls (§3.1, F16) | +| Gain-rule specification sensitivity | not reported | not reported | not reported | not reported | tested family; sign table fixed | | Whole analysis re-runs in a browser | some | no | no | no | endpoint parity verified | We claim no new reconstruction, no fitted *in vivo* potency, no replacement for regulatory toxicology and no live-animal result. The `live_lab` field of every notebook this pipeline produced is `null`. @@ -61,19 +60,19 @@ We claim no new reconstruction, no fitted *in vivo* potency, no replacement for The common failure this replaces is a pipeline that applies one transform to whatever number a source reported and gives the result one name. That erases two distinctions at once: a normalised functional response is not a fraction of occupied binding sites, and a row with no sourced number must not yield a small response at all. -Each row instead records the **parameter type** its source measured and an ordered **evidence distance** between that source and the modelled target: E0 (on-target); E1 (cross-species); E2 (related receptor); E3 (class extrapolation); E4 (unsupported). The admissible **transformation** is a total function of that pair (T20): a binding constant yields a physical occupancy only at E0 and a labelled proxy beyond it; a functional potency yields a direct engagement at E0 and a proxy beyond; anything at E4 is `not_modelled` and returns N/A. The result carries the name it earned, so a transferred number cannot be read as a measured one downstream. +Each row declares a **parameter type** and an ordered **evidence distance** between the stated source and modelled target: E0 (on-target); E1 (cross-species); E2 (related receptor); E3 (class extrapolation); E4 (unsupported). The admissible **transformation** is a total function of that pair (T20): a binding constant yields a physical occupancy only at E0 and a labelled proxy beyond it; a functional potency yields a direct engagement at E0 and a proxy beyond; anything at E4 is `not_modelled` and returns N/A. The result carries the name it earned, so a transferred number cannot be read as a measured one downstream. -Two properties make this a type discipline rather than a lookup table, and both are what we claim as the computing contribution. The refusal is **enforced**: requesting a value from an inadmissible row raises rather than returning one. And we state and check a **soundness invariant** over the whole pipeline — *no numeric engagement reaching a readout may originate from a row whose (parameter type, distance) pair does not admit one* — exhaustively, over every row and every entry point, with a companion check that dropping the inadmissible rows changes no downstream value, and a mutation check that makes the test fail with the right message when the rule table is altered. +Two properties distinguish this type discipline from a lookup table. First, requesting a value from an inadmissible row raises an error instead of returning a number. Second, tests check the invariant *no numeric engagement reaching an enumerated readout path may originate from a row whose (parameter type, distance) pair does not admit one*. The checks cover every shipped row and tested entry point; a companion test verifies that omitting inadmissible rows leaves the tested gain vectors unchanged, and a mutation test verifies that altering the rule table is detected. They do not check the primary assay record against each admitted numerical value, so source-to-value verification remains incomplete. ### 2.2 The extracts, and what the runtime does to them -Two committed extracts of MaleCNS v1.0 are used, streamed out of the ~1.1 GB public weight matrix on a CI runner and each under about 1 MB of JSON: **`named`**, a 1-hop neighbourhood of two seed cells (1126 nodes, 1360 edges), and **`taste_motor`**, the same seeds plus sensory types and an induced closure (1841 nodes, 19 066 edges). §3.2 adds a ladder of nested extracts of 1k, 5k, 10k, 25k, 50k cells grown by one fixed rule. +Two committed extracts of MaleCNS v1.0 are used, streamed out of the ~1.1 GB public weight matrix on a CI runner and each under about 1 MB of JSON: **`named`**, a 1-hop neighbourhood of two seed cells (1126 nodes, 1360 edges), and **`taste_motor`**, the same seeds plus sensory types and an induced closure (1841 nodes, 19 066 edges). A saved exploratory scale record describes nested extracts of 1k, 5k, 10k, 25k, 50k cells; its larger input cuts are absent from this review package. -**The extracts are structurally very different, and the difference turns out to be load-bearing (T19).** `named` has mean degree 1.208: 85% of its edges terminate on the 4 seed cells, only 184 of 1126 nodes receive any input, and only 25% of edges leave a node that itself receives input. It is an in-star, not a circuit. `taste_motor` has mean degree 10.4, 10% of edges onto seeds and 97% of edges leaving a node with input. Composition differs too, and one class is absent from both (0 cells), so a perturbation acting only through it cannot move either readout for substrate rather than domain reasons; the full census is in T19 and §S9(v). +**The extracts are structurally very different, and the difference turns out to be load-bearing (T19).** `named` has mean degree 1.208: 85% of its edges terminate on the 4 seed cells, only 184 of 1126 nodes receive any input, and only 25% of edges leave a node that itself receives input. It is dominated by edges entering seeds, with limited recurrent connectivity within this truncated cut. `taste_motor` has mean degree 10.4, 10% of edges onto seeds and 97% of edges leaving a node with input. Composition differs too, and one class is absent from both (0 cells), so a perturbation acting only through it cannot move either readout for substrate rather than domain reasons; the full census is in T19 and §S9(v). Two runtimes consume a gain patch: a deterministic clipped leaky rate iteration (~50 ms per evaluation of `named`) and a current-based LIF network [7] whose synaptic scale and background drive are tuned constants, not fits (§S3). -**The rate runtime row-normalises its weight matrix, and previous versions of this paper did not say so.** Each row of the signed matrix is divided by `max(Σ|W|, 1)` before any gain is applied, and gains then act as a presynaptic column scale, so every node's recurrent input is a *composition-weighted average of its presynaptic gains*. That builds part of the answer to RQ2 into the readout itself — an example of the general hazard that a runtime convention can carry a result — so it is documented here, in T2 and §S3, and swept in §3.3 against `none` and a degree-corrected denominator. The sign table the runtime iterates is published as T21: 5 of seven magnitudes (glutamate -0.4, histamine -0.5, dopamine +0.2, serotonin +0.2, octopamine +0.2) are asserted, varied by no member of the specification family and not bounded by the label-permuting null. +**The rate runtime row-normalises its weight matrix.** Each row of the signed matrix is divided by `max(Σ|W|, 1)` before any gain is applied, and gains then act as a presynaptic column scale. The resulting recurrent input depends on the weighted composition of a cell's presynaptic sources. This runtime convention can affect RQ2's classification; §3.3 compares it with no normalisation and a degree-corrected denominator. The sign table is reported in T21: 5 of seven magnitudes (glutamate -0.4, histamine -0.5, dopamine +0.2, serotonin +0.2, octopamine +0.2) are asserted and fixed in the gain-rule family and Sobol analysis. The transmitter-label null changes label assignments, not these magnitudes. ### 2.3 Input-dependence testing (RQ2) @@ -87,93 +86,95 @@ The instrument we consider most transferable asks, for one `compound × concentr | `sign_permute_weight_matched` | + the real weights on the real edges, and each transmitter's share of total outgoing weight | | the MaleCNS cut | everything | -The rank-3 rung changed in this revision. Because a gain patch acts *through* transmitter identity, a null that moves the weighted balance moves the drug's target set and the sign matrix at once, so the plain label permutation is relabelled a **joint target-set-and-sign null** and replaced by a constrained shuffle holding every tracked transmitter's share of outgoing weight within 0.002 of the real graph's. Permuting the node transmitter-label array preserves the transmitter histogram *by count* but not *by weight*: out-strength is heterogeneous, so on the `named` cut the cholinergic share of total outgoing synaptic weight is 0.619 on the real graph against 0.544 ± 0.027 across 1000 permutations, putting the real graph at the 100th percentile of its own null (T23). +Because a gain patch acts *through* transmitter identity, the plain label-permutation null changes both the target set and the sign assignment. It is therefore analysed separately as a **joint target-set-and-sign null**. The rank-3 rung uses a constrained shuffle holding every tracked transmitter's share of outgoing weight within 0.002 of the real graph's. Permuting labels preserves the transmitter histogram *by count* but not *by weight*: out-strength is heterogeneous, so on the `named` cut the cholinergic share of total outgoing synaptic weight is 0.619 on the real graph against 0.544 ± 0.027 across 1000 permutations (T23). -**Non-rejection is not equivalence, and the software no longer lets the prose say it is.** Each mode receives one of three verdicts against a margin prespecified as 5% of the vehicle readout (0.332 Hz on `named`): `distinguishable` when the empirical two-sided permutation probability `(k+1)/(n+1)` is at or below α; `equivalent_within_tolerance` when the test does not reject *and* the gap from the null median is below the margin; `indeterminate` otherwise. From the verdicts follow a **class** and a **necessary information level** — the weakest graph model the test could not distinguish from the real cut, never a claim that it reproduces the effect. Where a weaker model survives while a richer one is rejected the ladder is non-monotone, which the software reports and §3.2 counts. +Each mode receives one of three verdicts against a margin prespecified as 5% of the vehicle readout (0.332 rate-model units on `named`): `distinguishable` when the empirical two-sided permutation probability `(k+1)/(n+1)` is at or below α; `equivalent_within_tolerance` when the test does not reject *and* the observed point gap from the null median is below the margin; `indeterminate` otherwise. The point-gap verdict is descriptive, not a formal equivalence test with a confidence bound. A **class** and **necessary information level** summarise these component verdicts. The level identifies the weakest graph model this test could not distinguish from the real cut; it does not establish that the model reproduces the effect. Non-monotone ladders are flagged and counted in §3.2. -Three further conventions are prespecified and §S2 states them in full: a *relative* effect floor, a separation of **confirmatory** from **exploratory** tests with Benjamini-Hochberg applied across the latter, and a check of how many tests sit at the resolution floor, so a corrected count of zero is never mistaken for a finding. +Three further conventions are specified in §S2: a *relative* effect floor, separate planned profiles and exploratory landscapes, and a check of permutation resolution. Benjamini-Hochberg adjusts the exploratory component-test probabilities within the declared family; it does not provide a false-discovery guarantee for the derived cell classes. ### 2.4 The ablation ladder and the specification family (RQ3) *Which layer carries the information?* Four models predict the same cell with one layer denied: **A** receptor engagement alone; **B** mechanism gains on the graph's transmitter composition only, no edges; **C** the real cut and runtime with one generic multiplier; **D** the full model. They carry different units, so they are compared by how well each reproduces the full model's *ordering* of the library, scored on the **signed** Spearman ρ because a perfect inversion reproduces nothing. -Level C's multiplier changed in this revision: it could previously only *depress*, so no entry could come out positive and the level could not order a library half of whose entries raise the readout. Its magnitude is unchanged and its **sign** now comes from the mechanism table — one bit per compound — with the old rule retained as a floor (§S1). Levels B and D are not independent models, both being functions of the same gain vector, so §3.3 reports a **matched reference distribution** saying what agreement would have been for structural reasons alone. +Level C uses a generic magnitude with a mechanism-derived sign, one bit per compound. A depression-only generic rule is retained as a sensitivity reference (§S1). Levels B and D are not independent models, both being functions of the same gain vector, so §3.3 reports a **matched reference distribution** saying what agreement would have been for structural reasons alone. -*Which conclusions are properties of the data rather than of a modelling choice?* The parameter-to-gain mapping is the least defensible part of the pipeline — a textbook shape, but an asserted functional form. We prespecify 25 admissible specifications (five shapes at five coefficient scales, §S1.1) and re-derive every principal conclusion under every member; undecidable cells count against the conclusion and are reported separately, and the fraction is a fraction of *that family*, not a posterior probability. The topology conclusions are decided at 100 permutations per specification (resolution 0.010), corrected across the run's 100 structural tests. The two thresholds are likewise conventions, so the verdict is recomputed over the full 25%, 40%, 50% × 10%, 20%, 30% grid on 10 compounds. +*Which model-output predicates depend on the gain rule?* The parameter-to-gain mapping is the least defensible part of the pipeline — a textbook shape, but an asserted functional form. We evaluate 25 gain-rule specifications (five shapes at five coefficient scales, §S1.1) and re-derive each stated model-output predicate under every member; undecidable cells count against retention and are reported separately, and the fraction is a fraction of *that family*, not a posterior probability. The topology predicates are decided at 100 permutations per specification (resolution 0.010), corrected across the run's 100 structural tests. The two thresholds are likewise conventions, so the verdict is recomputed over the full 25%, 40%, 50% × 10%, 20%, 30% grid on 10 compounds. ### 2.5 Evaluation taxonomy -Agreement with published work is worth little unless its kind is named, so we separate three. **Implementation consistency** — the same inputs give the same outputs across transports, runtimes and runs (§3.7). **Literature concordance** — the model's *ordering* agrees with a published ordering, scored per entry and never as a bare mean, since an entry ordering two items can only return ±1; where the published ordering shares a source with the library, agreement is internal consistency, and the pipeline computes that overlap from DOIs and PMIDs rather than asserting it. **Independent out-of-sample validation** — against data entering neither the library nor the model, of which there is **none** at the circuit level. The ground-truth experiment of §3.1 validates the *instrument*, not the domain model, and we are careful not to let the first stand in for the second. +Agreement with published work is worth little unless its kind is named, so we separate three. **Implementation consistency** — the same inputs give the same outputs across transports, runtimes and runs (§3.7). **Literature concordance** — the model's *ordering* agrees with a published ordering, scored per entry and never as a bare mean, since an entry ordering two items can only return ±1; where the published ordering shares a source with the library, agreement is internal consistency, and the pipeline computes that overlap from DOIs and PMIDs rather than asserting it. **Independent out-of-sample validation** — against data entering neither the library nor the model, of which there is **none** at the circuit level. The planted-cycle experiment of §3.1 checks recovery for one synthetic mechanism, not the domain model. ### 2.6 Global uncertainty and the value of information -A one-at-a-time tornado cannot separate a factor's own effect from its interactions, so uncertainty is attributed with a variance-based (Sobol') analysis over every assumption the implementation exposes — potency, Hill coefficient, the engagement-to-gain transformation and its coefficient, the external drive, the cut's synapse-weight threshold, the transmitter-sign convention, the expression-weighting layer — plus a deliberately null factor, the spiking RNG seed, which the deterministic rate engine ignores. The design is Saltelli cross-sampling with Jansen estimators [71], [72], validated against the analytic Ishigami benchmark [73] in the same run. +A one-at-a-time sensitivity analysis cannot separate a factor's own effect from interactions, so uncertainty is attributed with a variance-based (Sobol') analysis over the selected factors: potency, Hill coefficient, engagement-to-gain transformation and coefficient, external drive, the cut's synapse-weight threshold, transmitter-label reassignment probability, expression weighting and a deliberately null spiking RNG seed, which the deterministic rate engine ignores. Transmitter-label reassignment does not vary the fixed sign magnitudes in T21. The design uses Saltelli cross-sampling with Jansen estimators [70], [71] and is checked against the analytic Ishigami benchmark [72]. -Estimates are left unclipped, so a factor whose true index is zero can be estimated below zero. A first-order index cannot be negative, so the **largest magnitude among the negative estimates is a measured lower bound on the estimator's error**, and that — not the null factor's own draw — is the noise floor we report. The budget becomes a ranking by `VOI_j = max(0, S_j) · Var(Y)` [76] — variance of a model output under assumed input ranges, not an expected gain in accuracy about a living fly. +Estimates are left unclipped, so a factor whose true index is zero can be estimated below zero. A first-order index cannot be negative, so the **largest magnitude among the negative estimates is a measured lower bound on the estimator's error**, and that — not the null factor's own draw — is the noise floor we report. For prioritisation we define a heuristic score `H_j = max(0, S_j) · Var(Y)`, conceptually motivated by information-value reasoning [75]. It ranks model-output variance under assumed ranges, not formal expected information value or biological accuracy. ## 3. Results -This section evaluates the instruments, not the domain. All values are regenerated by `scripts/reproduce_paper.py` and stored keyed in `papers/results.json`; the parameter library and the domain readouts they are demonstrated on are described in §3.6 and §S6. +This section evaluates the instruments. Core analysis values are regenerated by `scripts/reproduce_paper.py` and stored keyed in `papers/results.json`; larger scale-study values are imported from a saved record. The parameter library and the domain readouts they are demonstrated on are described in §3.6 and §S6. -### 3.1 Is the dependence instrument sound? Ground truth, power and convergence (Fig. F16, Tables T11, T22) +### 3.1 Synthetic positive controls and permutation convergence (Fig. F16, Tables T11, T22) -A negative from a dependence test is uninterpretable unless the test can be shown to return a positive when one exists, so the instrument is characterised against **planted ground truth** before it is used on anything. A recurrent cycle of known strength is planted in a synthetic network of the same size, density and composition as the real extract, so that the perturbation removes an amplification existing only while the cycle is intact: the *contrast*, not merely the level, depends on the wiring by construction. Strength 0 plants nothing and is the negative control. +The dependence procedure is checked against a **planted synthetic contrast**. A recurrent cycle of known strength is planted in a synthetic network of 250 nodes and about 1500 background edges. Its transmitter composition was chosen close to the `named` cut, but size and connectivity were not matched. The perturbation removes amplification that exists while the planted cycle is intact, so the treatment contrast depends on its wiring by construction. Strength 0 plants nothing and is the negative control. This design tests recovery for the planted mechanism, not power for all network effects. -The ladder recovered 2 of 3 planted effects at *n* = 200 with no false positive on the control. Over a power grid (6 independently generated networks per cell, permutation counts 50, 200, 1000) the detection rate by planted strength is strength 0.25: 0.00-0.17; strength 0.5: 0.67-0.83; strength 1: 1.00-1.00 — full power at strength 1, and at or below one in four at 0.25, so the instrument is **blind to weak planted structure and must not be read as ruling it out**. The empirical false-positive rate on the unplanted control is 0 of 18. Averaged over strengths, detection by permutation count is n=50: 0.61, n=200: 0.61, n=1000: 0.67: **effect size dominates, and permutation count buys little beyond a few hundred draws.** That is a design rule for this class of test — budget spent on permutations past the point of convergence is wasted, and it is the reason the scaling study of §3.2 can trade permutations for network size at all. +The ladder recovered 2 of 3 planted effects at *n* = 200 with no false positive on the control. In a small grid of 6 independently generated networks per cell and permutation counts 50, 200, 1000, detection by planted strength was strength 0.25: 0.00-0.17; strength 0.5: 0.67-0.83; strength 1: 1.00-1.00; the empirical false-positive count was 0 of 18. Detection was limited at 0.25. The count was n=50: 0.61, n=200: 0.61, n=1000: 0.67 across permutation budgets. The observed full-detection strengths are 1 in this finite grid. These synthetic results motivate a sensitivity warning for weak effects; they do not establish general power or justify a universal permutation budget. The convergence sweep (T11) separates the two quantities that permutation budget buys: a *verdict* at α = 0.05 settles from *n* = 50–400, while a quotable mid-range probability needs 25–400 draws to reach ±0.02 of its final value. The sweep re-reads prefixes of the same draws, so every row is a valid smaller sample and no extra runtime was spent producing it. -### 3.2 The verdict inverts with the extract, and recurrence predicts it (Figs. F6, F12; Tables T6, T12, T12b, T23, T27) +### 3.2 Dependence labels differ between the two extracts (Figs. F6, F12; Tables T6, T12, T12b, T23, T27) -**The confirmatory profiles** (full per-mode table: T6, §S2.1). At 1000 permutations per mode (resolution 0.0010), imidacloprid's effect on the sparse `named` extract is -6.173 Hz and the cell is **composition-dominated**: against the prespecified margin of 0.332 Hz it is equivalent within tolerance for rewire_degree_preserving, sign_permute_weight_matched, weight_permute, merely indeterminate for sign_permute, and distinguishable only from the Erdős–Rényi control (*p* = 9.99e-04 against 0.472 for degree-preserving rewiring). Fipronil on the same extract is the contrast — 0.900 Hz, **topology-dependent**, *p* = 0.006 and 0.007 for weight permutation and rewiring, necessary level `wiring_with_weighted_transmitter_balance` — so the instrument separates two predictions on one extract. Correcting the transmitter null moves no probability but changes what may be claimed: imidacloprid's gap from the null median falls from 2.652 Hz to 0.236 Hz, turning an indeterminate result into an equivalence one. +**The planned profiles** (full per-mode table: T6, §S2.1). At 1000 permutations per mode (resolution 0.0010), imidacloprid's effect on the sparse `named` extract is -6.173 rate-model units and the cell is **composition-dominated**: against the declared margin of 0.332 rate-model units, rewire_degree_preserving, sign_permute_weight_matched, weight_permute null comparisons pass the descriptive point-gap rule, sign_permute remain unresolved, and the Erdős–Rényi control is distinguishable (*p* = 9.99e-04 versus 0.472 for degree-preserving rewiring). Fipronil on the same extract is the contrast — 0.900 rate-model units, **topology-dependent**, *p* = 0.006 and 0.007 for weight permutation and rewiring, reported level `wiring_with_weighted_transmitter_balance`. The weight-matched transmitter null reduces imidacloprid's gap from 2.652 to 0.236 rate-model units, changing an unresolved comparison to a point-gap pass without establishing statistical equivalence. -**The landscape inverts between the two extracts.** Over 21 compounds × 4 concentrations = 84 cells at 1000 permutations, with Benjamini-Hochberg across the 168 structural tests and the relative effect floor applied, `named` gives **11 topology-dependent, 42 composition-dominated, 31 no-effect, 0 mixed** (uncorrected 14 / 39 / 31; on the absolute floor alone 16 / 55 / 13, the 18 cells the relative floor removes named in T12). Repeated unchanged on the denser `taste_motor` extract at 300 permutations, the same 84 cells give **54 / 0 / 30**: every cell above the floor is topology-dependent, imidacloprid included (*p* = 0.003, adjusted 0.006, class topology-dependent). +**Descriptive landscape.** The `named` cut has 85% of edges entering four seed cells. Over 21 compounds × 4 concentrations = 84 cells at 1000 permutations, with Benjamini-Hochberg across the 168 structural component tests and the relative effect floor applied, it gives **0 topology-dependent, 44 composition-dominated, 40 no-effect, 0 mixed** labels (uncorrected 11 / 33 / 40; on the absolute floor alone 0 / 59 / 25, with 15 cells removed by the relative floor in T12). The denser `taste_motor` extract uses 300 permutations and gives **44 / 0 / 40** among the same 84 cells; these classifications have no cell-level FDR guarantee and are conditional on each cut's nulls and runtime. -**Recurrence, not size, predicts the inversion.** Repeating the same profile across a ladder of nested extracts grown from one seed set by a single fixed rule (7 extracts in all, scale_1k, named, taste_motor, scale_5k, scale_10k, scale_25k, scale_50k; T27) settles the question of what the reversal tracks. The composition-dominated verdict appears on exactly 1 of them, `named`, and the inversion does not track the number of cells: `scale_1k` has 1000 nodes against `named`'s 1126 and a mean degree of 22.9 against 1.21, and it is already topology-dependent. Imidacloprid is topology-dependent on every rung but the in-star, and fipronil on all seven (per-rung sequences in §S2.4, full table T27). Both compounds' verdicts are settled on the clean ladder by the two-largest-rungs criterion of §S2.4, and at the denser middle rungs (scale_5k, scale_10k) the prediction is distinguishable from **every** null, its necessary level being the real network itself. The relative effect size is stable over this range (0.52-0.93 of the vehicle readout), so the change of verdict is not a change of what is being measured. +**Exploratory scale record.** The saved scale study reports 7 cuts (scale_1k, named, taste_motor, scale_5k, scale_10k, scale_25k, scale_50k; T27). The composition-dominated label appears on 1 (named), with per-rung sequences in §S2.4. The larger 5k–50k input cuts are not in this package, so those rungs cannot be independently rebuilt from the supplied data. The observed association of cut construction, density and label does not isolate recurrence as a cause; the two committed cuts also differ in composition and extraction rule. The lower permutation budgets on larger cuts (20-1000 draws) further limit comparisons. -Two honesty constraints travel with that result. The permutation budget must fall as the network grows (20-1000 draws, resolution 0.0010-0.0476), and at scale_10k, scale_25k, scale_50k the permutation budget falls to n = 20, where the smallest attainable probability is 0.048 against alpha = 0.05: a rejection there is the smallest the test can express — so the top rungs confirm the settled verdict rather than establishing it. And the two hops-limited cuts are built by a different rule from the nested rungs, so only comparisons *within* the ladder separate scale from construction. +The scale record uses decreasing permutation budgets as edge count grows (20-1000 draws; resolution 0.0010-0.0476). Comparisons between the hops-limited committed cuts and nested ranked-BFS scale cuts confound size with extraction rule. Results from the missing larger inputs are reported as imported observations, not reproduced outcomes. -**What is withdrawn, and what replaces it.** A degree-preserving rewire of an in-star is close to the identity, so the null had nothing to destroy; an extract with genuine recurrence does, and destroying it changes the answer. We withdraw the generalisation that most of what such a model predicts is carried by composition rather than topology, and the domain claim that went with it. What replaces it is stronger and methodological: **a dependence verdict is a joint property of the prediction and the extract, the extract's recurrence predicts which way it goes, and a negative measured on a sparse extract is not evidence about the network it was cut from.** +**Interpretation.** The sparse `named` cut has a large incoming-star component, whereas `taste_motor` contains more edges among nodes receiving input (T19). Degree-preserving rewiring may therefore change less of the effective input pattern on `named`. This is one plausible explanation for the contrasting labels, not a causal test of recurrence. Dependence classifications must be reported with the extract and runtime used. -**Multiplicity, working.** As a prespecified confirmatory profile, fipronil at 1 µM stands comfortably (*p* = 0.006 at *n* = 1000). Inside the exploratory landscape the same cell's adjusted probability is 0.050 and it does **not** survive correction; its corrected class there is composition-dominated against an uncorrected topology-dependent. +**Multiplicity.** The planned fipronil profile at 1 µM has component-test *p* = 0.006 at *n* = 1000. In the exploratory landscape, a corresponding adjusted component-test value is 0.062, with corrected class composition-dominated versus topology-dependent before adjustment. Benjamini-Hochberg applies to the declared structural test family; the cell class itself has no separate FDR guarantee. -**Two things the class label does not say.** The ladder is **non-monotone** in 11 of 84 cells on `named` and 0 of 84 on `taste_motor`, because adjacent rungs destroy different things; the level therefore means "the cheapest degraded input this test cannot tell apart from the real one", never "everything above it is indistinguishable too". And `composition-dominated` reports a failure to distinguish, never a reproduction: across the `named` landscape's mode verdicts, 281 are equivalent within tolerance, 50 indeterminate and 89 distinguishable. +The planned fipronil profile and its exploratory landscape class therefore answer different testing questions; a nominal result in the profile does not override the adjusted landscape label. + +**What the class label does not say.** The ladder is **non-monotone** in 0 of 84 cells on `named` and 1 of 84 on `taste_motor`, because adjacent rungs destroy different things. The level is the lowest-ranked degradation that passes the declared test and point-gap rule, not a monotone information threshold. A `composition-dominated` label also does not mean the null reproduces the effect: across `named` mode verdicts, 316 pass the descriptive point-gap rule, 44 are unresolved and 60 are distinguishable. ### 3.3 Which pipeline layer carries the information, and what agreement would have been anyway (Fig. F13, Tables T13, T13b, T24, T25) -Receptor engagement alone orders the library **backwards** (signed ρ -0.506 to -0.046) and is now the worst level at every concentration — constant across the table: yes (10 nM: A_receptor_only (-0.046); 100 nM: A_receptor_only (-0.348); 1 µM: A_receptor_only (-0.431); 10 µM: A_receptor_only (-0.506)). Under the historical level-C rule the worst level was *not* constant (10 nM: C_topology_only; 100 nM: A_receptor_only; 1 µM: A_receptor_only; 10 µM: A_receptor_only), so the previous claim that the topology-only level is the worst of the four was true of one column and false of the others. +Under the current library, receptor engagement alone has signed rank correlation ρ -0.579 to 0.095 with the full-model ordering. The lowest-correlation level is yes across the reported concentrations (10 nM: A_receptor_only (0.095); 100 nM: A_receptor_only (-0.519); 1 µM: A_receptor_only (-0.564); 10 µM: A_receptor_only (-0.579)); under the depression-only level-C reference, the lowest level varies (10 nM: C_topology_only; 100 nM: A_receptor_only; 1 µM: A_receptor_only; 10 µM: A_receptor_only). This comparison is conditional on the library and generic rule. -**The ablation conclusion is withdrawn.** Level C's multiplier applied a depression-only factor to every transmitter, so every entry was at most zero and the level could not represent disinhibition. With a direction-aware rule — same generic magnitude, one bit of sign — its rank correlation with the full model at 1 µM rises from 0.242 to 0.961. The sentence "the connectome without the pharmacology is not a cheap substitute for the connectome with it" was a statement about a sign-broken baseline and is retracted. +**Ablation reference.** A depression-only generic multiplier cannot represent disinhibition. The direction-aware rule uses the same generic magnitude with a mechanism-derived sign, and its topology-only rank correlation at 1 µM is 0.968 versus 0.161 under the depression-only reference. The difference shows that the generic rule affects the rank comparison. -**The composition-versus-full correlation carries no compound-level information.** It is 0.988 at 1 µM and 0.989 after dropping the compounds on the 0.05 gain floor, so it is not a tie artefact. But levels B and D are both functions of the same gain vector, and the matched reference says what would have been unsurprising: pharmacology-free pseudo-compounds with the gain shape the shipped rules produce give a median of 0.847 over 50 draws (5th–95th percentile 0.676 to 0.953), and permuting the compound labels leaves the observed value identically 0.988, as it must, because the permutation moves the (B, D) pairs as a block. +**The composition-versus-full correlation does not by itself establish compound-level information.** It is 0.921 at 1 µM and 0.826 after dropping compounds on the 0.05 gain floor, so it is not solely a tie artefact. Levels B and D both use the same gain vector. Pharmacology-free pseudo-compounds matched to the shipped gain shapes give a reference median of 0.847 over 50 draws (5th–95th percentile 0.676 to 0.953). Permuting compound labels moves B–D pairs together and leaves the observed correlation 0.921. -**And it is not robust to the runtime.** The correlation is 0.988 under the shipped normalisation and 0.906 with none, but 0.666 under a degree-corrected denominator at 1 µM and -0.372 at 10 nM: it survives row_abs, none and not a degree correction, so it is a property of *this runtime's readout* as much as of the perturbation. A sign disagreement between the two levels moves it by at most 0.057 and so is not what produces the agreement (§S9(i)). +**Runtime normalisation changes the agreement.** The correlation is 0.921 under the shipped normalisation and 0.935 with none, but 0.551 under a degree-corrected denominator at 1 µM and -0.583 at 10 nM. Agreement holds under row normalisation and without normalisation but weakens under degree correction, so it depends on the runtime's normalisation. A sign disagreement between the two levels moves it by at most 0.086 and so is not what produces the agreement (§S9(i)). -### 3.4 Specification robustness: which conclusions survive the modelling choices (Fig. F14, Tables T14, T14b, T15) +### 3.4 Gain-rule sensitivity of model-output predicates (Fig. F14, Tables T14, T14b, T15) -**The previous 25/25 topology result was vacuous, and the failure mode is general.** Two faults compounded. The matrix was computed at 6 shuffles per specification, where the smallest attainable permutation probability is 1/7 = 0.143, so `p ≤ 0.05` was arithmetically unattainable and the predicate silently fell back on a Gaussian criterion. And the specification context rebound the transformation on the assay modules but not on the engine the permutation path resolves, so **all 25 specifications fed default values to their nulls and produced identical effects**. A robustness matrix can thus report perfect stability because its alternatives never reached the code under test — a bug no determinism check can catch, since the run is perfectly reproducible either way. +**Specification test conditions.** The topology predicates use 100 shuffles per specification (resolution 0.010), and the gain specification is passed to the null engine. These conditions allow the empirical test to reach α = 0.05 and ensure the intended mapping is evaluated. The reported retention fractions are conditional on this implementation and the declared 25-member family. -Recomputed at 100 shuffles per specification (resolution 0.010), with the specification pushed into the null engine and Benjamini-Hochberg applied across the 100 structural tests, both topology conclusions still hold 25/25 and 25/25. The nuance the old matrix could not show is that a retention rate for a conclusion shaped as a *failure to reject* is only as strong as the equivalence behind it: of C3's 25, 17 reach equivalence within the prespecified margin and 8 are merely indeterminate. Under the 25 specifications: +Across 25 gain-rule specifications, C3 and C4 are retained in 25 and 25 cases. These C1–C7 conclusions are predicates over model outputs and the chosen null procedure, not robust biological conclusions. The transmitter sign magnitudes and transmitter-defined assignment are fixed across the family. For C3, 17 of the retained cases meet the descriptive point-gap criterion and 8 are indeterminate. Benjamini-Hochberg adjusts the 100 component tests in the run; it does not guarantee a false-discovery rate for predicate retention. -* The domain's most quotable claim — that the agonist suppresses activity — is retained by only 15 and **reversed** by all 10 monotone specifications: the same compound at the same engagement takes the network from 6.631 to 10.1 Hz (+52%) instead of from 6.631 to 0.458 Hz, with `g_ach` at 1.396 against 0.050. The direction is a property of one asserted term, not of the integration. -* A buffering claim is retained 18 of 25 and reversed by none, its 7 failures being undecidable. The previous explanation — that the monotone rules buffer "at least as much, never less" — is **false**: over the 6 reversing specifications with a defined index the mean gap runs -0.199 to -1.734 log₁₀ units, 3 buffer *less* (linear@1.50x, saturating@1.25x, saturating@1.50x), and the aggregate is carried by the extreme — an aggregate standing in for a range it does not represent. -* The matching amplification claim is weakest at 19 of 25. Retained by every specification: C2_rdl_disinhibition, C3_imidacloprid_topology_not_distinguishable, C4_fipronil_topology_exceeds, C7_map_bitter_veto. +* The domain's most quotable claim — that the agonist suppresses activity — is retained by only 15 and **reversed** by all 10 monotone specifications: the same compound at the same engagement takes the network from 6.631 to 10.1 rate-model units (+52%) instead of from 6.631 to 0.458 rate-model units, with `g_ach` at 1.396 against 0.050. The direction is a property of one asserted term, not of the integration. +* The buffering predicate is retained in 18 of 25 cases and undecidable in 7. Across 6 monotone specifications with a defined index, the mean gap spans -0.199 to -1.734 log₁₀ units; 3 buffer less than the shipped rule (linear@1.50x, saturating@1.25x, saturating@1.50x). +* The matching amplification claim is weakest at 18 of 25. Retained by every specification: C2_rdl_disinhibition, C3_imidacloprid_topology_not_distinguishable, C4_fipronil_topology_exceeds, C7_map_bitter_veto. -The threshold grid makes the same point about conventions rather than transformations: the split is **not** stable (`stable = no`), buffering holding in 9 of 9 cells (-0.362 to -1.224 log₁₀ units) and amplification in 8 of 9, flipping at the strictest corner (circuit_frac=0.50,vert_limit=0.10, gap -0.075), so ranges are quoted and never the point estimate. Two tables that appear to disagree in fact score different sets — T15 the 10 compounds of two mechanism classes, T8 the whole library — so acetylcholine is scored in one and outside the other's scope. +The threshold grid makes the same point about conventions rather than transformations: the split is **not** stable (`stable = no`), buffering holding in 9 of 9 cells (-0.362 to -1.224 log₁₀ units) and amplification in 7 of 9, flipping at the strictest corner (circuit_frac=0.40,vert_limit=0.10, circuit_frac=0.50,vert_limit=0.10, gap -0.154), so ranges are quoted and never the point estimate. Two tables that appear to disagree in fact score different sets — T15 the 10 compounds of two mechanism classes, T8 the whole library — so acetylcholine is scored in one and outside the other's scope. ### 3.5 Variance attribution, and what to measure next (Fig. F15, Tables T16, T17) -The Sobol' analysis (11 264 model evaluations at n_base = 1024, Var(Y) = 3.149 Hz²) resolves 2 of 9 factors in the strict sense of a bootstrap first-order interval excluding zero: gain_transform, weight_threshold — the **engagement-to-gain transformation** (S1 0.410, ST 0.626) and the cut's **synapse-weight threshold** (S1 0.294, ST 0.542). Everything else is unresolved at this sample size (drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter), including the external drive, whose S1 of 0.019 sits below the estimator's noise floor of 0.046 on an interval of (-0.072, +0.112). The floor is set by gain_coef, not by the declared null factor `lif_seed` at -0.007; the previous version of this section quoted the null factor and then ranked a factor inside the resulting noise. Interactions carry 0.344 of the variance raw and 0.272 with negatives clipped. On the analytic benchmark the estimator returns 0.3163, 0.4495, 0.0072 against the exact 0.3139, 0.4424, 0.0000 at n_base = 16 384 (max error 0.007). +The Sobol' analysis uses 11 264 model evaluations at n_base = 1024 and yields Var(Y) = 3.149 in squared rate-model units. Bootstrap first-order intervals exclude zero for 2 of 9 factors (gain_transform, weight_threshold), including the engagement-to-gain transformation (S1 0.410, ST 0.626) and cut weight threshold (S1 0.294, ST 0.542). Other factors are unresolved at this sample size (drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter). The first-order estimate for external drive is 0.019 with interval (-0.072, +0.112), and the negative-estimate noise diagnostic is 0.046, driven by gain_coef rather than the deliberately null factor (estimate -0.007). Estimated interaction shares are 0.344 raw and 0.272 after clipping negative estimates. Ishigami estimates are 0.3163, 0.4495, 0.0072 versus analytic values 0.3139, 0.4424, 0.0000 (maximum absolute error 0.007). These numerical diagnostics do not validate the biological model. -The top factor is **gain_transform** (1.290 of 3.149 Hz²), resolved by *calibration of receptor engagement against synaptic gain* — reading the *shape* of the response-versus-engagement curve at a known synapse, which is precisely the assumption §3.4 shows the suppression conclusion turns on. The second, **weight_threshold** (0.925 Hz²), needs **no experiment at all**: it is resolved by re-analysing synapse-confidence strata of data already held. +The author-defined variance-priority heuristic ranks **gain_transform** first (1.290 of 3.149 squared rate-model units). A direct assay of the response-versus-engagement curve at a known synapse (calibration of receptor engagement against synaptic gain) could constrain this factor. **weight_threshold** ranks second (0.925 squared rate-model units) and could be constrained by re-analysis of synapse-confidence strata. These rankings are conditional on assumed factor ranges, not formal expected values of information or evidence of biological importance. ### 3.6 What the refusal rule actually does (Fig. F11, Tables T10, T20) -The type system changes results rather than documentation. Of 102 library rows, 56 carry no admissible evidence and return N/A instead of a small plausible number, and the admissible transformation is decided by the (parameter type, distance) pair rather than by the author: 1 row earns a physical binding model, 1 a labelled binding proxy, 9 a direct functional engagement and 35 a transferred one. The single physical row is imidacloprid at insect_nAChR_native_dmel (Kd 2e-09 M) [77]; the library's previous binding constant, imidacloprid at insect_nAChR_beta1 (Kd 8.3e-11 M), is demoted to a cross-species proxy — the same number under a name that says it was transferred. Removing that one row would leave a branch of the type system with no instances, which is the kind of fact a lookup table hides and an exhaustively checked invariant surfaces. The full census, the domain readouts the engagements feed, and the literature-concordance analysis are in §S6 and §S10. +The type system changes which rows enter numerical predictions. Of 102 library rows, 63 are withheld and return N/A. The admitted categories (T20) contain physical binding (1), binding proxy (1), direct functional engagement (6) and transferred functional engagement (31) rows, assigned by declared type and distance. The binding model row is imidacloprid at insect_nAChR_native_dmel (Kd 2e-09 M) [76]; imidacloprid at insect_nAChR_beta1 (Kd 8.3e-11 M) is a transferred proxy. These admission categories do not independently verify every numeric source-value link. A row-level audit found assay mismatches and demoted affected entries; remaining numerical entries require source-to-parameter verification (§S6). ### 3.7 Reproduction, determinism and cross-backend parity (Fig. F1) -One command regenerates every figure, table and number in this paper from committed artifacts, with no download and no access to the 1.1 GB weight matrix: +The reproduction command regenerates the main analysis and its figures and tables from committed artifacts without downloading the 1.1 GB weight matrix. The saved scale-study record includes larger input cuts absent from the current package: ```bash python -m pip install -e ".[dev,viz]" @@ -182,199 +183,203 @@ python scripts/reproduce_paper.py --fast # reduced permutation and replicate c python scripts/reproduce_paper.py --only dependence --outdir /tmp/x ``` -Beyond the paper, `flylab run` turns a declarative spec into a self-describing run directory whose manifest records the spec, code and library hashes, the seeds and two hashes per artifact; the same spec and seed reproduce it byte for byte (`docs/WORKFLOW.md`). `papers/results.json` carries every scalar quoted above with its value, rendered text, unit and producing step; two runs at one commit and seed produce identical values, runtimes excepted, and a test asserts it. Every numeric span in the manuscript is a substitution from that record, so a number cannot enter the prose unless the pipeline produced it, and the test suite asserts that every key either document quotes exists in the record, that the record did not come from a reduced-effort run, and — new here, because the previous record's shuffle count did not match the shipped default and nothing caught it — that **every statistical parameter in the record equals the shipped default**. +Beyond the paper, `flylab run` turns a declarative spec into a self-describing run directory whose manifest records the spec, code and library hashes, the seeds and two hashes per artifact; the same spec and seed reproduce it byte for byte (`docs/WORKFLOW.md`). `papers/results.json` carries every scalar quoted above with its value, rendered text, unit and producing step; two runs at one commit and seed produce identical values, runtimes excepted, and a test asserts it. Numeric placeholders in this template are drawn from that record. Tests check placeholder presence, the effort mode and recorded statistical defaults; those tests do not independently verify source-value attribution or reproduce the absent larger scale inputs. -The same wheel runs in a browser tab on Pyodide and is published as a static site, so these assays can be re-run by opening one link with nothing installed; `tests/test_browser_bridge.py` asserts **endpoint parity** — equal JSON, every readout, warning and provenance field, for a representative call to each route, which is route-level equality rather than a whole-analysis equality claim. MaleCNS v1.0 and FlyWire are CC-BY 4.0 from their consortia [5], [6] and must be cited before this software; licences are listed under Statements, and an archival Zenodo deposit will carry the DOI **[to be minted on acceptance]**. +The wheel also runs in a browser tab on Pyodide; `tests/test_browser_bridge.py` asserts **endpoint parity** — equal JSON, every readout, warning and provenance field, for a representative call to each route, which is route-level equality rather than a whole-analysis equality claim. The MaleCNS download page lists CC-BY; the FlyWire public release guidelines list CC BY-NC 4.0 [5], [6]. Both sources require attribution. The code release has no separate archival DOI. ## 4. Discussion -**What transfers.** *First*, the dependence verdict is a joint property of a prediction and the extract it is measured on, and the extract's recurrence predicts which way it goes. The same validated instrument, with no parameter changed, classifies 42 of 84 cells composition-dominated on a sparse extract (mean degree 1.208) and 0 of 84 on a denser one (mean degree 10.4), and across 7 extracts the composition-dominated verdict survives on 1 — the in-star — while an extract with *fewer* nodes and twenty times the mean degree is already topology-dependent. The practical rule generalises to any pipeline of this shape: **a dependence result must be reported with the structural statistics of the extract it was measured on, and a negative measured on a sparse extract is not evidence about the network it was cut from.** Because the structural statistic that predicts the flip is computable before any permutation is run (T19, T27), this is a cheap check, not a research programme. +**Dependence interpretation.** The same procedure classifies 44 of 84 cells as composition-dominated on `named` and 0 of 84 on `taste_motor`, under their declared floors and test families. This is evidence that a dependence label is conditional on the cut, its construction and the runtime. The imported scale record adds descriptive observations across 7 cuts, but missing large-cut inputs and changing permutation budgets prevent a separate reproduction of the complete scale ladder. The observed structural differences suggest hypotheses about recurrence; they do not identify recurrence as the cause of the label change. *Second*, typing the evidence changes results rather than documentation: refusing inadmissible rows removed a class of small, plausible, invented numbers no reader could have told apart from measurements, and grading a source's distance demoted a cross-species constant to a labelled proxy without changing its value. Stated as an invariant and checked exhaustively, this is a property of the pipeline rather than a convention its authors intend to follow. *Third*, two of the instruments were themselves broken in ways that flattered the conclusions, and both were found by pointing the instruments inward: a null that moved the quantity it claimed to preserve exactly, and a specification family whose alternatives never reached the engine its predicates ran on, so a uniform robustness rate was guaranteed by construction. Both failures share a signature worth naming — **a test whose result was determined by its own implementation rather than by the data** — and both were invisible to determinism checks, because a record can be perfectly reproducible and still be produced by a test that could not have returned another answer. The specification family and the variance decomposition then converge on one weak point: the direction of the headline effect is set by an asserted functional form, which is also the largest resolved factor in the variance budget and the top-ranked measurement. -**Encoded assumptions versus generated observations.** The separation is itself generated rather than asserted: an audit of the dependency chain behind one readout returns 9 links (T18), of which exactly 1 link is a measurement of the system being simulated, 1 reuses a measurement of something else, 4 are asserted modelling choices and 3 are arithmetic on the links above. A companion decomposition — a different decomposition, not a repartition of those links — reports 5 facts, 1 model inference and 5 unknowns. That one link in nine is a measurement of the simulated system is worth stating plainly for any pipeline of this shape. +**Encoded assumptions versus generated observations.** An audit of the dependency chain behind one readout returns 9 links (T18): 1 link measures the simulated system, 1 reuses a measurement of something else, 4 are asserted model choices and 3 are computational links. This is a trace for one readout, not a partition of biological evidence quality. -**Relation to adjacent work.** Our version of the [69], [70] result survives neither a change of extract (§3.2) nor a change of the runtime's normalisation (§3.3), and both models normalise or regionally average, so the parallel is at least as consistent with a shared modelling artefact as with a property of receptor-weighted network perturbation. That is itself a methodological observation: a result reproduced across two pipelines that share a normalisation convention has not been independently confirmed. +**Relation to adjacent work.** Similar model-output associations in receptor-map network models [68], [69] motivate comparison of normalisation and spatial aggregation. The current experiments do not establish a shared mechanism or a biological correspondence between those domains. **What the pipeline cannot support**, and these limits are themselves computing results about what such a pipeline can establish. It cannot predict a dose–response in an animal: with no pharmacokinetics available, the free concentration is an *input*, not a derived quantity. It cannot support a cross-species safety claim, nor a subunit-level claim where the sources used hybrid constructs. It cannot support a general statement about whether perturbation on networks of this kind needs topology — only a statement per extract with that extract's structure reported. And it cannot produce a live-animal result, which no code path may write. -**Five model-specific caveats** are given in full in §S9: an asserted sign table exempt from the specification family and unbounded by the label-permuting null; the load-bearing row normalisation; perturbations applied across transmitter-defined rather than receptor-defined synapses, which the expression layer bounds rather than corrects; hops-limited extracts, whose missing inhibition can flip the sign of any readout and whose cost §3.2 now quantifies; and agreement of direction only between the two runtimes. +**Model-specific caveats** are detailed in §S9: fixed asserted transmitter sign magnitudes, load-bearing row normalisation, perturbations applied across transmitter-defined rather than receptor-defined synapses, truncated extracts with missing inputs, and only qualitative direction agreement between the two runtimes. The gain-rule family does not vary the sign magnitudes or transmitter assignment model. -**Prospective, not pre-registered.** The hypotheses in §S5 are written before any experiment and labelled as software predictions with no live result; a mutable repository is not a registry, so they become pre-registered only once a tagged release is externally archived. Two of them predict a readout whose *sign* is a modelling choice over this paper's own specification family, and the supplement now says so beside them. +**Prospective hypotheses.** The hypotheses in §S5 are software predictions with no live result. They are not preregistered; a mutable repository is not a registry. The direction of two readouts changes within the tested specification family, as stated beside those hypotheses. ## 5. Conclusion -We set out to make a network simulation say which of its inputs its predictions depend on, and to evaluate that machinery rather than assert it. The method is four instruments — typed evidence propagation under a checked soundness invariant, input-dependence testing against an ordered ladder with an explicit equivalence margin and multiplicity control, an ablation ladder scored against a matched reference, and specification-family robustness with variance attribution — none specific to the domain we demonstrate them on. Pointed at its own results, the method establishes that the headline effect's direction is a property of an asserted functional form; that a layer-agreement score near 1 can be an algebraic consequence of the pipeline's own normalisation; and — the result we would most want another group to break — that **a dependence verdict obtained on one extract of a network can invert on another extract of the same network, with recurrence rather than size predicting which way**. A "this prediction does not need the structure" result is only as good as the extract it was measured on, and the statistic that says so is cheap to compute in advance. +This study implements typed admission of parameter rows, graph-degradation comparisons, ablation references and gain-rule sensitivity checks for two cuts of MaleCNS v1.0. It finds that some model-output directions depend on the asserted gain rule and that descriptive dependence labels differ between the two cuts. The source-value audit, missing inputs for larger saved scale rungs and absent circuit-level experimental validation limit biological interpretation. The graph-null and point-gap results are conditional on the declared extract, runtime, margins and component-test family. -What would change the picture is short and specific, and the variance budget rather than taste names it: a calibration of the parameter-to-gain transformation at a known synapse; a reconstruction-confidence re-analysis, which needs no new measurement; a full landscape on an extract that is neither an in-star nor hops-limited; and one live assay. Until then this is a description of a method, and of a tool that tests itself. +Next steps are direct calibration of the engagement-to-gain transformation, verification of the remaining source-to-parameter links, reconstruction-confidence analyses, reproducible larger-cut tests and an independent circuit-level assay. Until these are available, the results are a methodological case study and software-derived predictions. ## Statements -**AI assistance.** The author defined the research questions, methods and analyses; verified the source literature, checking each cited publication against the value attributed to it; executed and reviewed every analysis; and accepts full responsibility for the scientific content. AI assistance was used for software implementation and for drafting and editing prose. No data were generated, imputed, estimated or extrapolated by a language model; every number in the results, tables and figures was produced by the reproduction pipeline from committed artifacts and is keyed in `papers/results.json`. +**AI use.** OpenAI Codex assisted code implementation, computational reruns and checks, bibliography and source-value screening, figure preparation, and language editing. Claude Opus 5.5 and Grok supplied informal manuscript critiques. The author remains responsible for the methods, analyses, citations and final text. A row-level audit identified unresolved source-value links; these are reported as limitations. Numerical results are produced by the stated code and parameter table, not measured in living animals. **Animal research.** No animal experiments were performed and no live-animal data are reported. The `live_lab` field of every notebook produced by this pipeline is `null`. The protocols in `protocols/` are unexecuted drafts. -**Data and code availability.** Code: https://github.com/pinkysworld/FlyLab (MIT). Connectome data: official MaleCNS v1.0 and FlyWire releases, CC-BY 4.0, downloaded and not redistributed; derived cuts are committed under `data/derived/` with their citation strings. Literature datasets: `data/literature/`, each value carrying a DOI or PMID. +**Data and code availability.** Versioned analysis code and derived cuts are available at https://github.com/pinkysworld/FlyLab/tree/d0ce21d777000b38bb57d786c308a059c4430c95 (Git revision `d0ce21d777000b38bb57d786c308a059c4430c95`; MIT). The repository contains keyed results and the literature parameter library; source-to-value support remains incomplete for some numeric rows. The official MaleCNS v1.0 dataset (CC-BY) and FlyWire public release (CC BY-NC 4.0) were downloaded but are not redistributed. Literature datasets are under `data/literature/`. A separate archival DOI for the code has not been minted; the article DOI is assigned by the journal upon publication. + +**Funding.** This study received no funding. + +**Competing interests.** The author declares no competing interests. -**Competing interests.** None declared. **Funding.** None. +**Author contributions.** The sole author was responsible for conceptualization, methodology, software, formal analysis, interpretation and writing. + +**SDG alignment.** The audit method and reusable analysis code are relevant to SDG 9, particularly target 9.5 on scientific research and innovation. The study does not measure progress toward an SDG target. ## References -[1] J. Dempster, *RatCVS*, University of Strathclyde, Glasgow, UK. Simulated pithed-rat cardiovascular pharmacology teaching package. +[1] J. Dempster, *RatCVS: Rat Cardiovascular System Simulation*, GitHub repository. [Online]. Available: https://github.com/johndempster/RatCVS (accessed 23 Sep. 2026). -[2] S. Okabe *et al.*, "A web-based simulation of cardiovascular pharmacology for undergraduate teaching," *Eur. J. Pharmacol.*, vol. 997, art. 177618, 2025. +[2] S. Okabe, T. Goto, D. Hirayama and Y. Nishimura, "Virtual rat web: A versatile simulation tool for pharmacology education in a variety of settings," *Eur. J. Pharmacol.*, vol. 997, art. 177618, 2025. DOI 10.1016/j.ejphar.2025.177618. -[3] *PharmVR*: immersive virtual-reality pharmacology training, preprint, 2025. +[3] K. Thiyagarajan, "PharmVR—Immersive Virtual Reality based Animal Simulator Software," *Preprints.org*, concept paper, 2025. DOI 10.20944/preprints202507.0959.v1. Not peer reviewed. -[4] "Virtual simulation in pharmacology and physiology education: an evaluation," *J. Educ. Eval. Health Prof.*, 2020. +[4] C. C. Ezeala, "Integration of computer-simulated practical exercises into undergraduate medical pharmacology education at Mulungushi University, Zambia," *J. Educ. Eval. Health Prof.*, vol. 17, art. 8, 2020. DOI 10.3352/jeehp.2020.17.8. [5] S. Dorkenwald *et al.*, "Neuronal wiring diagram of an adult brain," *Nature*, vol. 634, pp. 124–138, 2024. DOI 10.1038/s41586-024-07558-y. -[6] HHMI Janelia FlyEM, University of Cambridge, MRC LMB and Google Research, *MaleCNS v1.0: a synapse-resolution connectome of the adult male* Drosophila *central nervous system*, 2026. CC-BY 4.0. - -[7] P. K. Shiu, G. R. Sterne, N. Spiller *et al.*, "A leaky integrate-and-fire computational model based on the connectome of the entire adult *Drosophila* brain reveals insights into sensorimotor processing," *Nature*, vol. 634, 2024. +[6] S. Berg *et al.*, "Sexual dimorphism in the complete *Drosophila* male central nervous system connectome," *Cell*, vol. 189, no. 18, pp. 5504–5526.e15, 2026. DOI 10.1016/j.cell.2026.08.015. Associated dataset: MaleCNS v1.0, https://male-cns.janelia.org/download/ (CC-BY 4.0). -[8] C. Bass *et al.*, "Mutation analysis and functional characterisation of insect nicotinic acetylcholine receptors," *Insect Biochem. Mol. Biol.*, vol. 36, pp. 86–96, 2006. +[7] P. K. Shiu, G. R. Sterne, N. Spiller *et al.*, "A *Drosophila* computational brain model reveals sensorimotor processing," *Nature*, vol. 634, pp. 210–219, 2024. DOI 10.1038/s41586-024-07763-9. -[9] H. Dederer *et al.*, "Selectivity of neonicotinoids at hybrid insect nicotinic acetylcholine receptors," *Insect Biochem. Mol. Biol.*, vol. 41, p. 51, 2011. +[8] H. Dederer, M. Werr and T. Ilg, "Differential sensitivity of *Ctenocephalides felis* and *Drosophila melanogaster* nicotinic acetylcholine receptor α1 and α2 subunits in recombinant hybrid receptors to nicotinoids and neonicotinoid insecticides," *Insect Biochem. Mol. Biol.*, vol. 41, no. 1, pp. 51–61, 2011. DOI 10.1016/j.ibmb.2010.09.012. -[10] FlyConnectome, *Neurotransmitter ground-truth repository for* Drosophila *connectomes*, 2024. +[9] FlyConnectome, *Drosophila Neurotransmitters: Neurotransmitter Ground Truth*, GitHub repository, 2024. [Online]. Available: https://github.com/flyconnectome/drosophila_neurotransmitters (accessed 23 Sep. 2026). -[11] P. Mineault, "Deconstructing viral fly simulations," technical note, 2026. +[10] P. Mineault, "Deconstructing viral fly sims," *The NeuroAI Archive*, 14 Sep. 2026. [Online]. Available: https://www.neuroai.science/p/are-flies-playing-beat-saber. Newsletter essay, not peer reviewed. -[12] Y. Deng *et al.*, "Chemoconnectomics: mapping chemical transmission in *Drosophila*," *Neuron*, vol. 101, pp. 876–893, 2019. +[11] B. Deng *et al.*, "Chemoconnectomics: Mapping Chemical Transmission in *Drosophila*," *Neuron*, vol. 101, no. 5, pp. 876–893.e4, 2019. DOI 10.1016/j.neuron.2019.01.045. -[13] C. Bass, A. M. Puinean, M. Andrews *et al.*, "Mutation of a nicotinic acetylcholine receptor β subunit is associated with resistance to neonicotinoid insecticides in the aphid *Myzus persicae*," *BMC Neurosci.*, vol. 12, art. 51, 2011. DOI 10.1186/1471-2202-12-51. PMID 21627790. +[12] C. Bass, A. M. Puinean, M. Andrews *et al.*, "Mutation of a nicotinic acetylcholine receptor β subunit is associated with resistance to neonicotinoid insecticides in the aphid *Myzus persicae*," *BMC Neurosci.*, vol. 12, art. 51, 2011. DOI 10.1186/1471-2202-12-51. PMID 21627790. -[14] R. H. ffrench-Constant, T. A. Rocheleau, J. C. Steichen and A. E. Chalmers, "A point mutation in a *Drosophila* GABA receptor confers insecticide resistance," *Nature*, vol. 363, pp. 449–451, 1993. DOI 10.1038/363449a0. +[13] R. H. ffrench-Constant, T. A. Rocheleau, J. C. Steichen and A. E. Chalmers, "A point mutation in a *Drosophila* GABA receptor confers insecticide resistance," *Nature*, vol. 363, pp. 449–451, 1993. DOI 10.1038/363449a0. -[15] W. T. Garrood, C. T. Zimmer, O. Gutbrod *et al.*, "Influence of the RDL A301S mutation in the brown planthopper *Nilaparvata lugens* on the activity of phenylpyrazole insecticides," *Pestic. Biochem. Physiol.*, vol. 142, pp. 1–8, 2017. DOI 10.1016/j.pestbp.2017.01.007. PMID 29107231. +[14] W. T. Garrood, C. T. Zimmer, O. Gutbrod *et al.*, "Influence of the RDL A301S mutation in the brown planthopper *Nilaparvata lugens* on the activity of phenylpyrazole insecticides," *Pestic. Biochem. Physiol.*, vol. 142, pp. 1–8, 2017. DOI 10.1016/j.pestbp.2017.01.007. PMID 29107231. -[16] A. M. Hosie, H. A. Baylis, S. D. Buckingham and D. B. Sattelle, "Actions of the insecticide fipronil on dieldrin-sensitive and -resistant GABA receptors of *Drosophila melanogaster*," *Br. J. Pharmacol.*, vol. 115, no. 6, pp. 909–912, 1995. DOI 10.1111/j.1476-5381.1995.tb15896.x. PMID 7582519. +[15] A. M. Hosie, H. A. Baylis, S. D. Buckingham and D. B. Sattelle, "Actions of the insecticide fipronil on dieldrin-sensitive and -resistant GABA receptors of *Drosophila melanogaster*," *Br. J. Pharmacol.*, vol. 115, no. 6, pp. 909–912, 1995. DOI 10.1111/j.1476-5381.1995.tb15896.x. PMID 7582519. -[17] A. M. Hosie and D. B. Sattelle, "Allosteric modulation of an expressed homo-oligomeric GABA-gated chloride channel of *Drosophila melanogaster*," *Br. J. Pharmacol.*, vol. 117, no. 6, pp. 1229–1237, 1996. DOI 10.1111/j.1476-5381.1996.tb16720.x. PMID 8882620. +[16] A. M. Hosie and D. B. Sattelle, "Allosteric modulation of an expressed homo-oligomeric GABA-gated chloride channel of *Drosophila melanogaster*," *Br. J. Pharmacol.*, vol. 117, no. 6, pp. 1229–1237, 1996. DOI 10.1111/j.1476-5381.1996.tb16720.x. PMID 8882620. -[18] D. H. Kwon, K. S. Yoon, J. M. Clark and S. H. Lee, "A point mutation in a glutamate-gated chloride channel confers abamectin resistance in the two-spotted spider mite, *Tetranychus urticae* Koch," *Insect Mol. Biol.*, vol. 19, pp. 583–591, 2010. DOI 10.1111/j.1365-2583.2010.01017.x. PMID 20522121. +[17] D. H. Kwon, K. S. Yoon, J. M. Clark and S. H. Lee, "A point mutation in a glutamate-gated chloride channel confers abamectin resistance in the two-spotted spider mite, *Tetranychus urticae* Koch," *Insect Mol. Biol.*, vol. 19, pp. 583–591, 2010. DOI 10.1111/j.1365-2583.2010.01017.x. PMID 20522121. -[19] G. Le Goff, A. Hamon, J.-B. Bergé and M. Amichot, "Resistance to fipronil in *Drosophila simulans*: influence of two point mutations in the RDL GABA receptor subunit," *J. Neurochem.*, vol. 92, pp. 1295–1305, 2005. DOI 10.1111/j.1471-4159.2004.02922.x. PMID 15748149. +[18] G. Le Goff, A. Hamon, J.-B. Bergé and M. Amichot, "Resistance to fipronil in *Drosophila simulans*: influence of two point mutations in the RDL GABA receptor subunit," *J. Neurochem.*, vol. 92, pp. 1295–1305, 2005. DOI 10.1111/j.1471-4159.2004.02922.x. PMID 15748149. -[20] P. Menozzi, M. A. Shi, A. Lougarre, Z. H. Tang and D. Fournier, "Mutations of acetylcholinesterase which confer insecticide resistance in *Drosophila melanogaster* populations," *BMC Evol. Biol.*, vol. 4, art. 4, 2004. DOI 10.1186/1471-2148-4-4. PMID 15018651. +[19] P. Menozzi, M. A. Shi, A. Lougarre, Z. H. Tang and D. Fournier, "Mutations of acetylcholinesterase which confer insecticide resistance in *Drosophila melanogaster* populations," *BMC Evol. Biol.*, vol. 4, art. 4, 2004. DOI 10.1186/1471-2148-4-4. PMID 15018651. -[21] A. Mutero, M. Pralavorio, J. M. Bride and D. Fournier, "Resistance-associated point mutations in insecticide-insensitive acetylcholinesterase," *Proc. Natl. Acad. Sci. USA*, vol. 91, no. 13, pp. 5922–5926, 1994. DOI 10.1073/pnas.91.13.5922. PMID 8016090. +[20] A. Mutero, M. Pralavorio, J. M. Bride and D. Fournier, "Resistance-associated point mutations in insecticide-insensitive acetylcholinesterase," *Proc. Natl. Acad. Sci. USA*, vol. 91, no. 13, pp. 5922–5926, 1994. DOI 10.1073/pnas.91.13.5922. PMID 8016090. -[22] P. N. R. Usherwood, H. Vais, B. P. S. Khambay, T. G. E. Davies and M. S. Williamson, "Sensitivity of the *Drosophila para* sodium channel to DDT is not lowered by the super-kdr mutation M918T on the IIS4–S5 linker that profoundly reduces sensitivity to permethrin and deltamethrin," *FEBS Lett.*, vol. 579, no. 28, pp. 6317–6325, 2005. DOI 10.1016/j.febslet.2005.09.096. PMID 16263118. +[21] P. N. R. Usherwood, H. Vais, B. P. S. Khambay, T. G. E. Davies and M. S. Williamson, "Sensitivity of the *Drosophila para* sodium channel to DDT is not lowered by the super-kdr mutation M918T on the IIS4–S5 linker that profoundly reduces sensitivity to permethrin and deltamethrin," *FEBS Lett.*, vol. 579, no. 28, pp. 6317–6325, 2005. DOI 10.1016/j.febslet.2005.09.096. PMID 16263118. -[23] H. Vais, M. S. Williamson, S. J. Goodson *et al.*, "Activation of *Drosophila* sodium channels promotes modification by deltamethrin: reductions in affinity caused by knock-down resistance mutations," *J. Gen. Physiol.*, vol. 115, no. 3, pp. 305–318, 2000. DOI 10.1085/jgp.115.3.305. PMID 10694259. +[22] H. Vais, M. S. Williamson, S. J. Goodson *et al.*, "Activation of *Drosophila* sodium channels promotes modification by deltamethrin: reductions in affinity caused by knock-down resistance mutations," *J. Gen. Physiol.*, vol. 115, no. 3, pp. 305–318, 2000. DOI 10.1085/jgp.115.3.305. PMID 10694259. -[24] X. Wang, A. M. Puinean, A. O. O'Reilly *et al.*, "Mutations on M3 helix of *Plutella xylostella* glutamate-gated chloride channel confer unequal resistance to abamectin by two different mechanisms," *Insect Biochem. Mol. Biol.*, vol. 86, pp. 50–57, 2017. DOI 10.1016/j.ibmb.2017.05.006. PMID 28576654. +[23] X. Wang, A. M. Puinean, A. O. O'Reilly *et al.*, "Mutations on M3 helix of *Plutella xylostella* glutamate-gated chloride channel confer unequal resistance to abamectin by two different mechanisms," *Insect Biochem. Mol. Biol.*, vol. 86, pp. 50–57, 2017. DOI 10.1016/j.ibmb.2017.05.006. PMID 28576654. -[25] H. G. Zhang, R. H. ffrench-Constant and M. B. Jackson, "A unique amino acid of the *Drosophila* GABA receptor with influence on drug sensitivity by two mechanisms," *J. Physiol.*, vol. 479, no. 1, pp. 65–75, 1994. DOI 10.1113/jphysiol.1994.sp020278. PMID 7527461. +[24] H. G. Zhang, R. H. ffrench-Constant and M. B. Jackson, "A unique amino acid of the *Drosophila* GABA receptor with influence on drug sensitivity by two mechanisms," *J. Physiol.*, vol. 479, no. 1, pp. 65–75, 1994. DOI 10.1113/jphysiol.1994.sp020278. PMID 7527461. -[26] A. Cartereau, C. Martin and S. H. Thany, "Neonicotinoid insecticides differently modulate acetylcholine-induced currents on mammalian α7 nicotinic acetylcholine receptors," *Br. J. Pharmacol.*, vol. 175, no. 11, pp. 1987–1998, 2018. DOI 10.1111/bph.14018. +[25] A. Cartereau, C. Martin and S. H. Thany, "Neonicotinoid insecticides differently modulate acetylcholine-induced currents on mammalian α7 nicotinic acetylcholine receptors," *Br. J. Pharmacol.*, vol. 175, no. 11, pp. 1987–1998, 2018. DOI 10.1111/bph.14018. -[27] L. M. Cole, R. A. Nicholson and J. E. Casida, "Action of phenylpyrazole insecticides at the GABA-gated chloride channel," *Pestic. Biochem. Physiol.*, vol. 46, no. 1, pp. 47–54, 1993. DOI 10.1006/pest.1993.1035. +[26] L. M. Cole, R. A. Nicholson and J. E. Casida, "Action of phenylpyrazole insecticides at the GABA-gated chloride channel," *Pestic. Biochem. Physiol.*, vol. 46, no. 1, pp. 47–54, 1993. DOI 10.1006/pest.1993.1035. -[28] G. R. Dawson, K. A. Wafford, A. Smith *et al.*, "Anticonvulsant and adverse effects of avermectin analogs in mice are mediated through the γ-aminobutyric acid A receptor," *J. Pharmacol. Exp. Ther.*, vol. 295, no. 3, pp. 1051–1060, 2000. PMID 11082440. +[27] G. R. Dawson, K. A. Wafford, A. Smith *et al.*, "Anticonvulsant and adverse effects of avermectin analogs in mice are mediated through the γ-aminobutyric acid A receptor," *J. Pharmacol. Exp. Ther.*, vol. 295, no. 3, pp. 1051–1060, 2000. PMID 11082440. DOI 10.1016/S0022-3565(24)39005-6. -[29] K. Lees, M. Musgaard, S. Suwanmanee *et al.*, "Actions of agonists, fipronil and ivermectin on the predominant in vivo splice and edit variant (RDLbd, I/V) of the *Drosophila* GABA receptor expressed in *Xenopus laevis* oocytes," *PLoS ONE*, vol. 9, no. 5, art. e97468, 2014. DOI 10.1371/journal.pone.0097468. +[28] K. Lees, M. Musgaard, S. Suwanmanee *et al.*, "Actions of agonists, fipronil and ivermectin on the predominant in vivo splice and edit variant (RDLbd, I/V) of the *Drosophila* GABA receptor expressed in *Xenopus laevis* oocytes," *PLoS ONE*, vol. 9, no. 5, art. e97468, 2014. DOI 10.1371/journal.pone.0097468. -[30] T. Perry, D. G. Heckel, J. A. McKenzie and P. Batterham, "Effects of mutations in *Drosophila* nicotinic acetylcholine receptor subunits on sensitivity to insecticides targeting nicotinic acetylcholine receptors," *Pestic. Biochem. Physiol.*, vol. 102, no. 1, pp. 56–60, 2012. DOI 10.1016/j.pestbp.2011.10.010. +[29] T. Perry, D. G. Heckel, J. A. McKenzie and P. Batterham, "Effects of mutations in *Drosophila* nicotinic acetylcholine receptor subunits on sensitivity to insecticides targeting nicotinic acetylcholine receptors," *Pestic. Biochem. Physiol.*, vol. 102, no. 1, pp. 56–60, 2012. DOI 10.1016/j.pestbp.2011.10.010. -[31] G. S. Ratra and J. E. Casida, "GABA receptor subunit composition relative to insecticide potency and selectivity," *Toxicol. Lett.*, vol. 122, no. 3, pp. 215–222, 2001. DOI 10.1016/s0378-4274(01)00366-6. PMID 11489356. +[30] G. S. Ratra and J. E. Casida, "GABA receptor subunit composition relative to insecticide potency and selectivity," *Toxicol. Lett.*, vol. 122, no. 3, pp. 215–222, 2001. DOI 10.1016/s0378-4274(01)00366-6. PMID 11489356. -[32] G. S. Ratra, S. G. Kamita and J. E. Casida, "Role of human GABA_A receptor β3 subunit in insecticide toxicity," *Toxicol. Appl. Pharmacol.*, vol. 172, no. 3, pp. 233–240, 2001. DOI 10.1006/taap.2001.9154. PMID 11312652. +[31] G. S. Ratra, S. G. Kamita and J. E. Casida, "Role of human GABA_A receptor β3 subunit in insecticide toxicity," *Toxicol. Appl. Pharmacol.*, vol. 172, no. 3, pp. 233–240, 2001. DOI 10.1006/taap.2001.9154. PMID 11312652. -[33] Q. Shan, J. L. Haddrill and J. W. Lynch, "Ivermectin, an unconventional agonist of the glycine receptor chloride channel," *J. Biol. Chem.*, vol. 276, no. 16, pp. 12556–12564, 2001. DOI 10.1074/jbc.M011264200. +[32] Q. Shan, J. L. Haddrill and J. W. Lynch, "Ivermectin, an unconventional agonist of the glycine receptor chloride channel," *J. Biol. Chem.*, vol. 276, no. 16, pp. 12556–12564, 2001. DOI 10.1074/jbc.M011264200. -[34] K. Tasman, S. A. Rands and J. J. L. Hodge, "The power of *Drosophila melanogaster* for modeling neonicotinoid effects on pollinators and identifying novel mechanisms," *Front. Physiol.*, vol. 12, art. 659440, 2021. DOI 10.3389/fphys.2021.659440. +[33] K. Tasman, S. A. Rands and J. J. L. Hodge, "The power of *Drosophila melanogaster* for modeling neonicotinoid effects on pollinators and identifying novel mechanisms," *Front. Physiol.*, vol. 12, art. 659440, 2021. DOI 10.3389/fphys.2021.659440. -[35] M. Tomizawa and J. E. Casida, "Selective toxicity of neonicotinoids attributable to specificity of insect and mammalian nicotinic receptors," *Annu. Rev. Entomol.*, vol. 48, pp. 339–364, 2003. DOI 10.1146/annurev.ento.48.091801.112731. PMID 12208819. +[34] M. Tomizawa and J. E. Casida, "Selective toxicity of neonicotinoids attributable to specificity of insect and mammalian nicotinic receptors," *Annu. Rev. Entomol.*, vol. 48, pp. 339–364, 2003. DOI 10.1146/annurev.ento.48.091801.112731. PMID 12208819. -[36] M. Tomizawa and J. E. Casida, "Neonicotinoid insecticide toxicology: mechanisms of selective action," *Annu. Rev. Pharmacol. Toxicol.*, vol. 45, pp. 247–268, 2005. DOI 10.1146/annurev.pharmtox.45.120403.095930. +[35] M. Tomizawa and J. E. Casida, "Neonicotinoid insecticide toxicology: mechanisms of selective action," *Annu. Rev. Pharmacol. Toxicol.*, vol. 45, pp. 247–268, 2005. DOI 10.1146/annurev.pharmtox.45.120403.095930. -[37] M. Tomizawa, D. L. Lee and J. E. Casida, "Neonicotinoid insecticides: molecular features conferring selectivity for insect versus mammalian nicotinic receptors," *J. Agric. Food Chem.*, vol. 48, no. 12, pp. 6016–6024, 2000. DOI 10.1021/jf000873c. PMID 11312774. +[36] M. Tomizawa, D. L. Lee and J. E. Casida, "Neonicotinoid insecticides: molecular features conferring selectivity for insect versus mammalian nicotinic receptors," *J. Agric. Food Chem.*, vol. 48, no. 12, pp. 6016–6024, 2000. DOI 10.1021/jf000873c. PMID 11312774. -[38] A. J. Wolstenholme, "Glutamate-gated chloride channels," *J. Biol. Chem.*, vol. 287, no. 48, pp. 40232–40238, 2012. DOI 10.1074/jbc.R112.406280. +[37] A. J. Wolstenholme, "Glutamate-gated chloride channels," *J. Biol. Chem.*, vol. 287, no. 48, pp. 40232–40238, 2012. DOI 10.1074/jbc.R112.406280. -[39] A. M. Allen, M. C. Neville, S. Birtles *et al.*, "A single-cell transcriptomic atlas of the adult *Drosophila* ventral nerve cord," *eLife*, vol. 9, art. e54074, 2020. DOI 10.7554/eLife.54074. PMID 32314735. +[38] A. M. Allen, M. C. Neville, S. Birtles *et al.*, "A single-cell transcriptomic atlas of the adult *Drosophila* ventral nerve cord," *eLife*, vol. 9, art. e54074, 2020. DOI 10.7554/eLife.54074. PMID 32314735. -[40] V. Croset, C. D. Treiber and S. Waddell, "Cellular diversity in the *Drosophila* midbrain revealed by single-cell transcriptomics," *eLife*, vol. 7, art. e34550, 2018. DOI 10.7554/eLife.34550. PMID 29671739. +[39] V. Croset, C. D. Treiber and S. Waddell, "Cellular diversity in the *Drosophila* midbrain revealed by single-cell transcriptomics," *eLife*, vol. 7, art. e34550, 2018. DOI 10.7554/eLife.34550. PMID 29671739. -[41] K. Davie, J. Janssens, D. Koldere *et al.*, "A single-cell transcriptome atlas of the aging *Drosophila* brain," *Cell*, vol. 174, no. 4, pp. 982–998.e20, 2018. DOI 10.1016/j.cell.2018.05.057. PMID 29909982. +[40] K. Davie, J. Janssens, D. Koldere *et al.*, "A single-cell transcriptome atlas of the aging *Drosophila* brain," *Cell*, vol. 174, no. 4, pp. 982–998.e20, 2018. DOI 10.1016/j.cell.2018.05.057. PMID 29909982. -[42] L. Enell, Y. Hamasaka, A. Kolodziejczyk and D. R. Nässel, "γ-Aminobutyric acid (GABA) signaling components in *Drosophila*: immunocytochemical localization of GABA_B receptors in relation to the GABA_A receptor subunit RDL and a vesicular GABA transporter," *J. Comp. Neurol.*, vol. 505, no. 1, pp. 18–31, 2007. DOI 10.1002/cne.21472. PMID 17729251. +[41] L. Enell, Y. Hamasaka, A. Kolodziejczyk and D. R. Nässel, "γ-Aminobutyric acid (GABA) signaling components in *Drosophila*: immunocytochemical localization of GABA_B receptors in relation to the GABA_A receptor subunit RDL and a vesicular GABA transporter," *J. Comp. Neurol.*, vol. 505, no. 1, pp. 18–31, 2007. DOI 10.1002/cne.21472. PMID 17729251. -[43] N. Konstantinides, K. Kapuralin, C. Fadil *et al.*, "Phenotypic convergence: distinct transcription factors regulate common terminal features," *Cell*, vol. 174, no. 3, pp. 622–635.e13, 2018. DOI 10.1016/j.cell.2018.05.021. PMID 29909983. +[42] N. Konstantinides, K. Kapuralin, C. Fadil *et al.*, "Phenotypic convergence: distinct transcription factors regulate common terminal features," *Cell*, vol. 174, no. 3, pp. 622–635.e13, 2018. DOI 10.1016/j.cell.2018.05.021. PMID 29909983. -[44] H. Li, J. Janssens, M. De Waegeneer *et al.*, "Fly Cell Atlas: a single-nucleus transcriptomic atlas of the adult fruit fly," *Science*, vol. 375, no. 6584, art. eabk2432, 2022. DOI 10.1126/science.abk2432. PMID 35239393. +[43] H. Li, J. Janssens, M. De Waegeneer *et al.*, "Fly Cell Atlas: a single-nucleus transcriptomic atlas of the adult fruit fly," *Science*, vol. 375, no. 6584, art. eabk2432, 2022. DOI 10.1126/science.abk2432. PMID 35239393. -[45] C. I. Bliss, "The toxicity of poisons applied jointly," *Ann. Appl. Biol.*, vol. 26, no. 3, pp. 585–615, 1939. DOI 10.1111/j.1744-7348.1939.tb06990.x. +[44] C. I. Bliss, "The toxicity of poisons applied jointly," *Ann. Appl. Biol.*, vol. 26, no. 3, pp. 585–615, 1939. DOI 10.1111/j.1744-7348.1939.tb06990.x. -[46] W. R. Greco, G. Bravo and J. C. Parsons, "The search for synergy: a critical review from a response surface perspective," *Pharmacol. Rev.*, vol. 47, no. 2, pp. 331–385, 1995. PMID 7568331. +[45] W. R. Greco, G. Bravo and J. C. Parsons, "The search for synergy: a critical review from a response surface perspective," *Pharmacol. Rev.*, vol. 47, no. 2, pp. 331–385, 1995. PMID 7568331. -[47] Q. Huang, Y. Deng, T. Zhan and Y. He, "Synergistic and antagonistic effects of piperonyl butoxide in fipronil-susceptible and resistant rice stem borers, *Chilo suppressalis*," *J. Insect Sci.*, vol. 10, art. 182, 2010. DOI 10.1673/031.010.14142. PMID 21062143. +[46] Q. Huang, Y. Deng, T. Zhan and Y. He, "Synergistic and antagonistic effects of piperonyl butoxide in fipronil-susceptible and resistant rice stem borrers [sic], *Chilo suppressalis*," *J. Insect Sci.*, vol. 10, art. 182, 2010. DOI 10.1673/031.010.14142. PMID 21062143. -[48] S. Loewe and H. Muischnek, "Über Kombinationswirkungen: 1. Mitteilung: Hilfsmittel der Fragestellung," *Naunyn-Schmiedebergs Arch. Exp. Pathol. Pharmakol.*, vol. 114, pp. 313–326, 1926. DOI 10.1007/BF01952257. +[47] S. Loewe and H. Muischnek, "Über Kombinationswirkungen: 1. Mitteilung: Hilfsmittel der Fragestellung," *Naunyn-Schmiedebergs Arch. Exp. Pathol. Pharmakol.*, vol. 114, pp. 313–326, 1926. DOI 10.1007/BF01952257. -[49] S. Loewe, "The problem of synergism and antagonism of combined drugs," *Arzneimittelforschung*, vol. 3, no. 6, pp. 285–290, 1953. PMID 13081480. +[48] S. Loewe, "The problem of synergism and antagonism of combined drugs," *Arzneimittelforschung*, vol. 3, no. 6, pp. 285–290, 1953. PMID 13081480. -[50] D. Philippou, V. Borzatta, E. Capparella *et al.*, "The use of substituted alkynyl phenoxy derivatives of piperonyl butoxide to control insecticide-resistant pests," *Pest Manag. Sci.*, vol. 72, no. 10, pp. 1946–1950, 2016. DOI 10.1002/ps.4234. PMID 26800141. +[49] D. Philippou, V. Borzatta, E. Capparella *et al.*, "The use of substituted alkynyl phenoxy derivatives of piperonyl butoxide to control insecticide-resistant pests," *Pest Manag. Sci.*, vol. 72, no. 10, pp. 1946–1950, 2016. DOI 10.1002/ps.4234. PMID 26800141. -[51] Y. C. Zhu, J. Yao, J. Adamczyk and R. Luttrell, "Synergistic toxicity and physiological impact of imidacloprid alone and binary mixtures with seven representative pesticides on honey bee (*Apis mellifera*)," *PLoS ONE*, vol. 12, no. 5, art. e0176837, 2017. DOI 10.1371/journal.pone.0176837. PMID 28467462. +[50] Y. C. Zhu, J. Yao, J. Adamczyk and R. Luttrell, "Synergistic toxicity and physiological impact of imidacloprid alone and binary mixtures with seven representative pesticides on honey bee (*Apis mellifera*)," *PLoS ONE*, vol. 12, no. 5, art. e0176837, 2017. DOI 10.1371/journal.pone.0176837. PMID 28467462. -[52] P. J. Daborn, J. L. Yen, M. R. Bogwitz *et al.*, "A single P450 allele associated with insecticide resistance in *Drosophila*," *Science*, vol. 297, no. 5590, pp. 2253–2256, 2002. DOI 10.1126/science.1074170. PMID 12351787. +[51] P. J. Daborn, J. L. Yen, M. R. Bogwitz *et al.*, "A single P450 allele associated with insecticide resistance in *Drosophila*," *Science*, vol. 297, no. 5590, pp. 2253–2256, 2002. DOI 10.1126/science.1074170. PMID 12351787. -[53] D. G. Folk, C. Han and T. J. Bradley, "Water acquisition and partitioning in *Drosophila melanogaster*: effects of selection for desiccation-resistance," *J. Exp. Biol.*, vol. 204, no. 19, pp. 3323–3331, 2001. DOI 10.1242/jeb.204.19.3323. PMID 11606606. +[52] D. G. Folk, C. Han and T. J. Bradley, "Water acquisition and partitioning in *Drosophila melanogaster*: effects of selection for desiccation-resistance," *J. Exp. Biol.*, vol. 204, no. 19, pp. 3323–3331, 2001. DOI 10.1242/jeb.204.19.3323. PMID 11606606. -[54] R. Fusetto, S. Denecke, T. Perry, R. A. J. O'Hair and P. Batterham, "Partitioning the roles of CYP6G1 and gut microbes in the metabolism of the insecticide imidacloprid in *Drosophila melanogaster*," *Sci. Rep.*, vol. 7, art. 11339, 2017. DOI 10.1038/s41598-017-09800-2. PMID 28900131. +[53] R. Fusetto, S. Denecke, T. Perry, R. A. J. O'Hair and P. Batterham, "Partitioning the roles of CYP6G1 and gut microbes in the metabolism of the insecticide imidacloprid in *Drosophila melanogaster*," *Sci. Rep.*, vol. 7, art. 11339, 2017. DOI 10.1038/s41598-017-09800-2. PMID 28900131. -[55] N. Joußen, D. G. Heckel, M. Haas, I. Schuphan and B. Schmidt, "Metabolism of imidacloprid and DDT by P450 CYP6G1 expressed in cell cultures of *Nicotiana tabacum*," *Pest Manag. Sci.*, vol. 64, no. 1, pp. 65–73, 2008. DOI 10.1002/ps.1472. PMID 17912692. +[54] N. Joußen, D. G. Heckel, M. Haas, I. Schuphan and B. Schmidt, "Metabolism of imidacloprid and DDT by P450 CYP6G1 expressed in cell cultures of *Nicotiana tabacum* suggests detoxification of these insecticides in *Cyp6g1*-overexpressing strains of *Drosophila melanogaster*, leading to resistance," *Pest Manag. Sci.*, vol. 64, no. 1, pp. 65–73, 2008. DOI 10.1002/ps.1472. PMID 17912692. -[56] S. C. Piyankarage, H. Augustin, Y. Grosjean, D. E. Featherstone and S. A. Shippy, "Hemolymph amino acid analysis of individual *Drosophila* larvae," *Anal. Chem.*, vol. 80, no. 4, pp. 1201–1207, 2008. DOI 10.1021/ac701785z. +[55] S. C. Piyankarage, H. Augustin, Y. Grosjean, D. E. Featherstone and S. A. Shippy, "Hemolymph amino acid analysis of individual *Drosophila* larvae," *Anal. Chem.*, vol. 80, no. 4, pp. 1201–1207, 2008. DOI 10.1021/ac701785z. -[57] S. C. Piyankarage, D. E. Featherstone and S. A. Shippy, "Nanoliter hemolymph sampling and analysis of individual adult *Drosophila melanogaster*," *Anal. Chem.*, vol. 84, no. 10, pp. 4460–4466, 2012. DOI 10.1021/ac3002319. +[56] S. C. Piyankarage, D. E. Featherstone and S. A. Shippy, "Nanoliter hemolymph sampling and analysis of individual adult *Drosophila melanogaster*," *Anal. Chem.*, vol. 84, no. 10, pp. 4460–4466, 2012. DOI 10.1021/ac3002319. -[58] S. Suchail, L. Debrauwer and L. P. Belzunces, "Metabolism of imidacloprid in *Apis mellifera*," *Pest Manag. Sci.*, vol. 60, no. 3, pp. 291–296, 2004. DOI 10.1002/ps.772. PMID 15025241. +[57] S. Suchail, L. Debrauwer and L. P. Belzunces, "Metabolism of imidacloprid in *Apis mellifera*," *Pest Manag. Sci.*, vol. 60, no. 3, pp. 291–296, 2004. DOI 10.1002/ps.772. PMID 15025241. -[59] S. Suchail, G. De Sousa, R. Rahmani and L. P. Belzunces, "In vivo distribution and metabolisation of ¹⁴C-imidacloprid in different compartments of *Apis mellifera* L.," *Pest Manag. Sci.*, vol. 60, no. 11, pp. 1056–1062, 2004. DOI 10.1002/ps.895. +[58] S. Suchail, G. De Sousa, R. Rahmani and L. P. Belzunces, "In vivo distribution and metabolisation of ¹⁴C-imidacloprid in different compartments of *Apis mellifera* L.," *Pest Manag. Sci.*, vol. 60, no. 11, pp. 1056–1062, 2004. DOI 10.1002/ps.895. -[60] W. Cao, L. Song, J. Cheng *et al.*, "An automated rapid iterative negative geotaxis assay for analyzing adult climbing behavior in a *Drosophila* model of neurodegeneration," *J. Vis. Exp.*, no. 127, art. 56507, 2017. DOI 10.3791/56507. PMID 28931001. +[59] W. Cao, L. Song, J. Cheng *et al.*, "An automated rapid iterative negative geotaxis assay for analyzing adult climbing behavior in a *Drosophila* model of neurodegeneration," *J. Vis. Exp.*, no. 127, art. 56507, 2017. DOI 10.3791/56507. PMID 28931001. -[61] M. Ö. Çevik and A. Erden, "The course of habituation of the proboscis extension reflex can be predicted by sucrose responsiveness in *Drosophila*," *PLoS ONE*, vol. 7, no. 6, art. e39863, 2012. DOI 10.1371/journal.pone.0039863. PMID 22761915. +[60] M. Ö. Çevik and A. Erden, "The course of habituation of the proboscis extension reflex can be predicted by sucrose responsiveness in *Drosophila*," *PLoS ONE*, vol. 7, no. 6, art. e39863, 2012. DOI 10.1371/journal.pone.0039863. PMID 22761915. -[62] W. Chen, X. Gu, Y. T. Yang, P. Batterham and T. Perry, "Dual nicotinic acetylcholine receptor subunit gene knockouts reveal limits to functional redundancy," *Pestic. Biochem. Physiol.*, vol. 184, art. 105118, 2022. DOI 10.1016/j.pestbp.2022.105118. PMID 35715057. +[61] W. Chen, X. Gu, Y. T. Yang, P. Batterham and T. Perry, "Dual nicotinic acetylcholine receptor subunit gene knockouts reveal limits to functional redundancy," *Pestic. Biochem. Physiol.*, vol. 184, art. 105118, 2022. DOI 10.1016/j.pestbp.2022.105118. PMID 35715057. -[63] A. S. French, M. J. Sellier, M. Ali Agha *et al.*, "Dual mechanism for bitter avoidance in *Drosophila*," *J. Neurosci.*, vol. 35, no. 9, pp. 3990–4004, 2015. DOI 10.1523/JNEUROSCI.1312-14.2015. PMID 25740527. +[62] A. S. French, M. J. Sellier, M. Ali Agha *et al.*, "Dual mechanism for bitter avoidance in *Drosophila*," *J. Neurosci.*, vol. 35, no. 9, pp. 3990–4004, 2015. DOI 10.1523/JNEUROSCI.1312-14.2015. PMID 25740527. -[64] J. W. Gargano, I. Martin, P. Bhandari and M. S. Grotewiel, "Rapid iterative negative geotaxis (RING): a new method for assessing age-related locomotor decline in *Drosophila*," *Exp. Gerontol.*, vol. 40, no. 5, pp. 386–395, 2005. DOI 10.1016/j.exger.2005.02.005. PMID 15919590. +[63] J. W. Gargano, I. Martin, P. Bhandari and M. S. Grotewiel, "Rapid iterative negative geotaxis (RING): a new method for assessing age-related locomotor decline in *Drosophila*," *Exp. Gerontol.*, vol. 40, no. 5, pp. 386–395, 2005. DOI 10.1016/j.exger.2005.02.005. PMID 15919590. -[65] H. Y. Lee, L. Zhou, S. Ghanta *et al.*, "Mechanisms of naturally evolved ethanol resistance in *Drosophila melanogaster*," *J. Exp. Biol.*, 2014. DOI 10.1242/jeb.110510. +[64] J. D. Fry, "Mechanisms of naturally evolved ethanol resistance in *Drosophila melanogaster*," *J. Exp. Biol.*, vol. 217, no. 22, pp. 3996–4003, 2014. DOI 10.1242/jeb.110510. -[66] F. Martelli, Z. Zhongyuan, J. Wang *et al.*, "Low doses of the neonicotinoid insecticide imidacloprid induce ROS triggering neurological and metabolic impairments in *Drosophila*," *Proc. Natl. Acad. Sci. USA*, vol. 117, no. 41, pp. 25840–25850, 2020. DOI 10.1073/pnas.2011828117. PMID 32989137. +[65] F. Martelli, Z. Zhongyuan, J. Wang *et al.*, "Low doses of the neonicotinoid insecticide imidacloprid induce ROS triggering neurological and metabolic impairments in *Drosophila*," *Proc. Natl. Acad. Sci. USA*, vol. 117, no. 41, pp. 25840–25850, 2020. DOI 10.1073/pnas.2011828117. PMID 32989137. -[67] M. J. Burton, I. R. Mellor, I. R. Duce *et al.*, "Differential resistance of insect sodium channels with kdr mutations to deltamethrin, permethrin and DDT," *Insect Biochem. Mol. Biol.*, vol. 41, no. 9, pp. 723–732, 2011. DOI 10.1016/j.ibmb.2011.05.004. +[66] M. J. Burton, I. R. Mellor, I. R. Duce *et al.*, "Differential resistance of insect sodium channels with kdr mutations to deltamethrin, permethrin and DDT," *Insect Biochem. Mol. Biol.*, vol. 41, no. 9, pp. 723–732, 2011. DOI 10.1016/j.ibmb.2011.05.004. -[68] A. A. Lazar, T. Liu, M. K. Turkcan and Y. Zhou, "Accelerating with FlyBrainLab the discovery of the functional logic of the *Drosophila* brain in the connectomic and synaptomic era," *eLife*, vol. 10, art. e62362, 2021. DOI 10.7554/eLife.62362. PMID 33616035. +[67] A. A. Lazar, T. Liu, M. K. Turkcan and Y. Zhou, "Accelerating with FlyBrainLab the discovery of the functional logic of the *Drosophila* brain in the connectomic and synaptomic era," *eLife*, vol. 10, art. e62362, 2021. DOI 10.7554/eLife.62362. PMID 33616035. -[69] I. Mindlin, R. Herzog, L. Belloli *et al.*, "Whole brain modelling for simulating pharmacological interventions on patients with disorders of consciousness," *Commun. Biol.*, vol. 7, art. 1176, 2024. DOI 10.1038/s42003-024-06852-9. PMID 39300281. +[68] I. Mindlin, R. Herzog, L. Belloli *et al.*, "Whole brain modelling for simulating pharmacological interventions on patients with disorders of consciousness," *Commun. Biol.*, vol. 7, art. 1176, 2024. DOI 10.1038/s42003-024-06852-9. PMID 39300281. -[70] G. Deco, J. Cruzat, J. Cabral *et al.*, "Whole-brain multimodal neuroimaging model using serotonin receptor maps explains non-linear functional effects of LSD," *Curr. Biol.*, vol. 28, no. 19, pp. 3065–3074.e6, 2018. DOI 10.1016/j.cub.2018.07.083. PMID 30270185. +[69] G. Deco, J. Cruzat, J. Cabral *et al.*, "Whole-brain multimodal neuroimaging model using serotonin receptor maps explains non-linear functional effects of LSD," *Curr. Biol.*, vol. 28, no. 19, pp. 3065–3074.e6, 2018. DOI 10.1016/j.cub.2018.07.083. PMID 30270185. -[71] A. Saltelli, P. Annoni, I. Azzini, F. Campolongo, M. Ratto and S. Tarantola, "Variance based sensitivity analysis of model output: design and estimator for the total sensitivity index," *Comput. Phys. Commun.*, vol. 181, no. 2, pp. 259–270, 2010. DOI 10.1016/j.cpc.2009.09.018. +[70] A. Saltelli, P. Annoni, I. Azzini, F. Campolongo, M. Ratto and S. Tarantola, "Variance based sensitivity analysis of model output: design and estimator for the total sensitivity index," *Comput. Phys. Commun.*, vol. 181, no. 2, pp. 259–270, 2010. DOI 10.1016/j.cpc.2009.09.018. -[72] M. J. W. Jansen, "Analysis of variance designs for model output," *Comput. Phys. Commun.*, vol. 117, no. 1–2, pp. 35–43, 1999. DOI 10.1016/S0010-4655(98)00154-4. +[71] M. J. W. Jansen, "Analysis of variance designs for model output," *Comput. Phys. Commun.*, vol. 117, no. 1–2, pp. 35–43, 1999. DOI 10.1016/S0010-4655(98)00154-4. -[73] T. Ishigami and T. Homma, "An importance quantification technique in uncertainty analysis for computer models," in *Proc. 1st Int. Symp. Uncertainty Modeling and Analysis (ISUMA)*, 1990, pp. 398–403. DOI 10.1109/ISUMA.1990.151285. +[72] T. Ishigami and T. Homma, "An importance quantification technique in uncertainty analysis for computer models," in *Proc. 1st Int. Symp. Uncertainty Modeling and Analysis (ISUMA)*, 1990, pp. 398–403. DOI 10.1109/ISUMA.1990.151285. -[74] S. Steegen, F. Tuerlinckx, A. Gelman and W. Vanpaemel, "Increasing transparency through a multiverse analysis," *Perspect. Psychol. Sci.*, vol. 11, no. 5, pp. 702–712, 2016. DOI 10.1177/1745691616658637. PMID 27694465. +[73] S. Steegen, F. Tuerlinckx, A. Gelman and W. Vanpaemel, "Increasing transparency through a multiverse analysis," *Perspect. Psychol. Sci.*, vol. 11, no. 5, pp. 702–712, 2016. DOI 10.1177/1745691616658637. PMID 27694465. -[75] U. Simonsohn, J. P. Simmons and L. D. Nelson, "Specification curve analysis," *Nat. Hum. Behav.*, vol. 4, pp. 1208–1214, 2020. DOI 10.1038/s41562-020-0912-z. PMID 32719546. +[74] U. Simonsohn, J. P. Simmons and L. D. Nelson, "Specification curve analysis," *Nat. Hum. Behav.*, vol. 4, pp. 1208–1214, 2020. DOI 10.1038/s41562-020-0912-z. PMID 32719546. -[76] R. A. Howard, "Information value theory," *IEEE Trans. Syst. Sci. Cybern.*, vol. 2, no. 1, pp. 22–26, 1966. DOI 10.1109/TSSC.1966.300074. +[75] R. A. Howard, "Information value theory," *IEEE Trans. Syst. Sci. Cybern.*, vol. 2, no. 1, pp. 22–26, 1966. DOI 10.1109/TSSC.1966.300074. -[77] M. Tomizawa, B. Latli and J. E. Casida, "Novel neonicotinoid-agarose affinity column for *Drosophila* and *Musca* nicotinic acetylcholine receptors," *J. Neurochem.*, vol. 67, no. 4, pp. 1669–1676, 1996. DOI 10.1046/j.1471-4159.1996.67041669.x. PMID 8858952. +[76] M. Tomizawa, B. Latli and J. E. Casida, "Novel neonicotinoid-agarose affinity column for *Drosophila* and *Musca* nicotinic acetylcholine receptors," *J. Neurochem.*, vol. 67, no. 4, pp. 1669–1676, 1996. DOI 10.1046/j.1471-4159.1996.67041669.x. PMID 8858952. --- diff --git a/papers/IJRC_FlyLab_draft.md.in b/papers/IJRC_FlyLab_draft.md.in index bf8950d..b830581 100644 --- a/papers/IJRC_FlyLab_draft.md.in +++ b/papers/IJRC_FlyLab_draft.md.in @@ -9,7 +9,7 @@ the same way from `papers/SUPPLEMENT.md.in`. --> # Which Inputs Does a Network Simulation Depend On? Typed Evidence Propagation and Input-Dependence Testing, Evaluated on Cuts of a Synapse-Resolution *Drosophila* Connectome -**Research article, submitted draft for the International Journal of Research in Computing (IJRC)** +**Research article draft for the International Journal of Research in Computing (IJRC)** Software: https://github.com/pinkysworld/FlyLab Version {{flylab_version}}, commit `{{git_sha}}`, compound library SHA-256 `{{library_sha256}}`, map `{{map_id}}`. @@ -18,48 +18,47 @@ Version {{flylab_version}}, commit `{{git_sha}}`, compound library SHA-256 `{{li ## Abstract -A simulation built on a measured network and parameterised from published literature cannot, on its own, say which of its inputs its predictions depend on. We present a method that makes that question routine and testable, and an open implementation, FlyLab, that answers it per prediction. The method has four parts: **typed evidence propagation**, in which the parameter a source measured and its graded distance from the modelled target jointly decide which transformation is admissible, enforced by a soundness invariant checked exhaustively over the parameter library ({{ev_rows_not_modelled}} of {{ev_rows}} rows refused rather than approximated); **input-dependence testing** against an ordered ladder of degraded copies of the network, reporting — with a prespecified equivalence margin and multiplicity control — the weakest input the test cannot distinguish from the real one; an **ablation ladder** over pipeline layers scored against a matched reference distribution; and **specification-family robustness** with variance attribution that ranks what to measure next. We evaluate the method rather than the domain: the dependence instrument is validated against planted ground truth with a power surface, its permutation budget is characterised by a convergence sweep, and the analysis is deterministic across two backends. **The principal result is methodological: a dependence verdict obtained on one extract of a network can invert on another extract of the same network, and the extract's recurrence, not its size, predicts which way.** Demonstrated on cuts of the public MaleCNS v1.0 *Drosophila* connectome, a prediction is composition-dominated on a sparse 1-hop extract (mean degree {{named_mean_degree}}, {{named_share_onto_seeds}} of edges onto four seed cells) and topology-dependent on every extract with real recurrence, including one with *fewer* nodes; across nested extracts of {{scale_ladder_sizes}} cells the verdict settles as topology-dependent. Such a verdict is therefore a joint property of the prediction and the extract, and must be reported with the extract's structure. The central limitation is unchanged: the pipeline evaluates a model's behaviour, and its parameter-to-gain rules are asserted rather than measured. +Literature-parameterised network simulations can report predictions without showing which inputs carry them. We evaluate a method combining typed evidence gates, graph-degradation nulls, a matched ablation reference, gain-rule specifications and variance attribution on two cuts of the MaleCNS v1.0 *Drosophila* connectome. The current 21-compound library has {{ev_rows}} parameter rows; {{ev_rows_not_modelled}} do not yield numerical engagement after unsupported sources are withheld. This is a syntactic admission rule, not proof that every admitted assay value is correct: source-to-parameter verification remains incomplete. The incoming-star-like `named` cut has {{named_share_onto_seeds}} of edges entering four seed cells. Under the declared floors and adjusted component tests, its {{dep_land_cells}} compound–concentration cells have {{dep_land_topology_dependent}} topology-dependent and {{dep_land_composition_dominated}} composition-dominated descriptive labels, while the denser `taste_motor` cut has {{dep_land_taste_topology_dependent}} and {{dep_land_taste_composition_dominated}}. This contrast illustrates extract sensitivity, not a causal effect of recurrence. The gain transformation controls the direction of one model-output predicate. The larger scale study is imported from a saved record, and no independent circuit-level biological validation is available. -**Index terms:** computational method, scientific software, provenance, evidence typing, network simulation, null models, equivalence testing, multiple comparisons, specification robustness, sensitivity analysis, reproducible research, connectomics. +**Index terms:** computational method, scientific software, provenance, evidence typing, network simulation, null models, permutation testing, multiple comparisons, specification sensitivity, reproducible research, connectomics. ## 1. Introduction -A large class of scientific software has the same shape: a measured network, a table of parameters compiled from published literature, a rule that turns those parameters into perturbations of the network, and a numerical runtime that produces a prediction. Power grids, metabolic models, epidemic contact networks and connectome simulations are all built this way. Such a pipeline readily produces a number; what it does not produce is an answer to the question a reviewer asks first — *which of its inputs does this number actually depend on?* Without that answer the prediction cannot be audited, because a result carried entirely by an asserted coefficient is indistinguishable, at the output, from one carried by the measured structure. +A large class of scientific software has the same shape: a measured network, a table of parameters compiled from literature and declared model assumptions, a rule that turns those parameters into perturbations of the network, and a numerical runtime that produces a prediction. Related pipelines occur in several scientific domains; this study tests one connectome application. Such a pipeline readily produces a number; what it does not produce is an answer to the question a reviewer asks first — *which of its inputs does this number actually depend on?* Without that answer the prediction cannot be audited, because a result carried entirely by an asserted coefficient is indistinguishable, at the output, from one carried by the measured structure. Three properties make the question hard to answer in practice. The parameters are heterogeneous: sources measure different quantities, in different preparations, and are often simply absent, so a pipeline that coerces them into one numeric field silently manufactures evidence. The network enters through a runtime whose own conventions — normalisation, sign, iteration — are modelling choices that can carry part of the answer. And the usual test of whether structure matters is a permutation against a shuffled network, whose *non-rejection* is routinely written up as though the shuffle had reproduced the effect. -This paper contributes a method addressing all three (§2), an implementation that runs it per prediction rather than once per paper, and — the part we think is missing from this literature — an evaluation of the instruments themselves (§3). We demonstrate the method on cuts of a public synapse-resolution *Drosophila* connectome, a domain supplying exactly the hard case: a measured network, a literature parameter table with large gaps, and asserted transformation rules. The headline finding is about the method rather than the domain: **a dependence verdict obtained on one extract of a network can invert on another extract of the same network.** +This paper contributes a method addressing all three (§2), an implementation that runs it per prediction rather than once per paper, and — the part we think is missing from this literature — an evaluation of the instruments themselves (§3). We demonstrate the method on cuts of a public synapse-resolution *Drosophila* connectome, a domain supplying exactly the hard case: a measured network, a literature parameter table with large gaps, and asserted transformation rules. The central contribution is an audit of evidence admission, graph-null preservation and specification propagation. The two cuts illustrate how a descriptive dependence label changes with extraction. -We organise the work around four research questions. +We organise the scientific evaluation around three research questions; reproducibility is engineering support. * **RQ1.** Can heterogeneous published parameters be turned into computational perturbations *without* erasing their provenance or their parameter semantics, and can that discipline be stated as a checkable invariant rather than a convention? * **RQ2.** For a given prediction, which inputs does it depend on — and is that answer a property of the prediction, or of the extract of the network it was measured on? * **RQ3.** Which conclusions survive alternative admissible specifications, and which assumptions dominate the output's variance? -* **RQ4.** Can the entire analysis be reproduced identically, across backends, from auditable artifacts? -The answers are: yes, by grading the evidence, refusing invalid transformations and checking the refusal exhaustively; **it is a property of the extract**, which is the paper's main result; not all of them, and an undocumented normalisation inside the runtime carries part of RQ2's answer; and yes. An earlier version of this work answered RQ2 with "most predictions do not need the network", measured on a single extract — a conclusion this revision withdraws and explains. +The results show a finite, checkable admission rule for RQ1; extract- and runtime-dependent labels for RQ2; and specification-dependent model-output predicates for RQ3. These findings are conditional on the current parameter table and selected cuts. The RQ1 tests check declared types, readout paths and gain vectors; they do not verify the primary assay value behind each admitted row. ### 1.1 Related work and what is new here -**Network runtimes without parameterised perturbation.** Teaching simulators [1]–[4] model dose–response with no network at all. At the other extreme, FlyWire [5] and MaleCNS v1.0 [6] publish netlists with per-node annotations [10], [12]; Shiu *et al.* [7] show a literature-parameterised model over such a netlist reproduces a sensorimotor control, and downstream ports repeat the exercise with varying rigour [11]. FlyBrainLab [68] builds executable circuits directly from connectome data, but a perturbation, where present, enters as a hand-set multiplier with nothing behind it — exactly the case typed propagation is meant to replace. +**Network runtimes without parameterised perturbation.** RatCVS [1] and Virtual Rat Web [2] lack connectomes; PharmVR [3] is a preprint; [4] evaluates teaching. At the other extreme, FlyWire [5] and MaleCNS v1.0 [6] publish netlists with per-node annotations [9], [11]; Shiu *et al.* [7] show a literature-parameterised model over such a netlist reproduces a sensorimotor control, and downstream ports repeat the exercise with varying rigour [10]. FlyBrainLab [67] builds executable circuits from connectome data but does not describe the per-row pharmacological admission rule evaluated here. -**Parameter-map-informed network models.** On human structural connectomes, intervention is simulated by scaling regional responses with PET density maps [69], [70] — the closest published relative, whose most relevant result is also a negative one: in Mindlin *et al.* [69] the simulated improvement tracked the *mean* density of the activated receptors rather than where they sit. Methodologically, §2.4 is a multiverse analysis [74], [75] applied to a simulation, and §2.6 uses Saltelli cross-sampling [71] with Jansen estimators [72] validated on Ishigami [73] and ranked by value of information [76]. +**Parameter-map-informed network models.** On human structural connectomes, intervention is simulated by scaling regional responses with PET density maps [68], [69]. In Mindlin *et al.* [68], maximum simulated perturbation distance correlated with mean normalised receptor density, whereas their tested receptor-map/node-strength divergence metric did not. Methodologically, §2.4 is a multiverse analysis [73], [74] applied to a simulation, and §2.6 uses Saltelli cross-sampling [70] with Jansen estimators [71] validated on Ishigami [72] and ranked by an author-defined variance-priority heuristic motivated by information-value theory [75]. -**The domain evidence base.** The parameter library, and the supporting datasets for resistance alleles, expression, mixtures, exposure and behavioural endpoints, are compiled from the insect-versus-vertebrate pharmacology literature [8], [9], [13]–[67] and documented in §S4 and §S6; what matters methodologically is that it is typical of such tables — incomplete, heterogeneous in what each source measured, and internally inconsistent with its own citations in ten places that `data/literature/README.md` records without editing the library to match. Those gaps are the reason the type system changes results rather than documentation. +**The domain evidence base.** The parameter library combines published measurements, transferred values and explicit model assumptions [8], [12]–[66]. A row-level source audit found assay mismatches and unsupported numeric values; corrected rows are withheld from numerical engagement, while other exact values remain under primary-source review. The type system enforces declared categories but cannot itself establish that a paper supports the entered potency. -**Position.** Table I states what is new narrowly. The transferable part is not the domain coupling but the instrumentation around it: routine, per-prediction, *validated* testing of which inputs a prediction depends on, with non-rejection separated from equivalence, multiplicity controlled, and the test's own power characterised against planted ground truth. +**Position.** Table I states what is new narrowly. The transferable part is not the domain coupling but the instrumentation around it: routine per-prediction diagnostics with non-rejection distinguished from a descriptive point-gap pass, adjusted component tests and limited positive-control experiments. These are not a cell-level false-discovery guarantee or a general power estimate. -**Table I. What each system provides.** +**Table I. Capabilities described in the cited implementations.** A negative entry means the cited work does not report the capability; it is not a claim about every implementation in the category. -| Capability | Teaching sims [1]–[4] | Connectome LIF runtimes [7], [11] | FlyBrainLab [68] | Receptor-map whole-brain models [69], [70] | FlyLab | +| Capability | Teaching simulators [1]–[4] | Connectome LIF model [7] | FlyBrainLab [67] | Regional receptor models [68], [69] | This study | |---|---|---|---|---|---| | Named cells at synapse resolution | no | yes | yes | no (regions) | yes | | Compound **and free concentration** as the input | yes (organ level) | no | no | partly (receptor-map scaling) | yes | | Typed evidence with per-row provenance and distance | no | no | no | no | yes | | Two species scored at the same dose | no | no | no | no | yes | -| Dependence of each prediction on its inputs, tested | no | rarely | no | partly [69] | yes, routinely | -| The dependence test itself validated against ground truth | n/a | no | no | no | yes (§3.1, F16) | -| Specification robustness of the stated conclusions | no | no | no | no | yes | +| Dependence of each prediction on its inputs, tested | not reported | not reported in [7] | not reported | partly [68] | yes, for declared nulls | +| Synthetic recovery cases evaluated | n/a | not assessed | not assessed | not assessed | limited positive controls (§3.1, F16) | +| Gain-rule specification sensitivity | not reported | not reported | not reported | not reported | tested family; sign table fixed | | Whole analysis re-runs in a browser | some | no | no | no | endpoint parity verified | We claim no new reconstruction, no fitted *in vivo* potency, no replacement for regulatory toxicology and no live-animal result. The `live_lab` field of every notebook this pipeline produced is `null`. @@ -70,19 +69,19 @@ We claim no new reconstruction, no fitted *in vivo* potency, no replacement for The common failure this replaces is a pipeline that applies one transform to whatever number a source reported and gives the result one name. That erases two distinctions at once: a normalised functional response is not a fraction of occupied binding sites, and a row with no sourced number must not yield a small response at all. -Each row instead records the **parameter type** its source measured and an ordered **evidence distance** between that source and the modelled target: {{ev_distance_labels}}. The admissible **transformation** is a total function of that pair (T20): a binding constant yields a physical occupancy only at E0 and a labelled proxy beyond it; a functional potency yields a direct engagement at E0 and a proxy beyond; anything at E4 is `not_modelled` and returns N/A. The result carries the name it earned, so a transferred number cannot be read as a measured one downstream. +Each row declares a **parameter type** and an ordered **evidence distance** between the stated source and modelled target: {{ev_distance_labels}}. The admissible **transformation** is a total function of that pair (T20): a binding constant yields a physical occupancy only at E0 and a labelled proxy beyond it; a functional potency yields a direct engagement at E0 and a proxy beyond; anything at E4 is `not_modelled` and returns N/A. The result carries the name it earned, so a transferred number cannot be read as a measured one downstream. -Two properties make this a type discipline rather than a lookup table, and both are what we claim as the computing contribution. The refusal is **enforced**: requesting a value from an inadmissible row raises rather than returning one. And we state and check a **soundness invariant** over the whole pipeline — *no numeric engagement reaching a readout may originate from a row whose (parameter type, distance) pair does not admit one* — exhaustively, over every row and every entry point, with a companion check that dropping the inadmissible rows changes no downstream value, and a mutation check that makes the test fail with the right message when the rule table is altered. +Two properties distinguish this type discipline from a lookup table. First, requesting a value from an inadmissible row raises an error instead of returning a number. Second, tests check the invariant *no numeric engagement reaching an enumerated readout path may originate from a row whose (parameter type, distance) pair does not admit one*. The checks cover every shipped row and tested entry point; a companion test verifies that omitting inadmissible rows leaves the tested gain vectors unchanged, and a mutation test verifies that altering the rule table is detected. They do not check the primary assay record against each admitted numerical value, so source-to-value verification remains incomplete. ### 2.2 The extracts, and what the runtime does to them -Two committed extracts of MaleCNS v1.0 are used, streamed out of the ~1.1 GB public weight matrix on a CI runner and each under about 1 MB of JSON: **`named`**, a 1-hop neighbourhood of two seed cells ({{named_nodes}} nodes, {{named_edges}} edges), and **`taste_motor`**, the same seeds plus sensory types and an induced closure ({{taste_nodes}} nodes, {{taste_edges}} edges). §3.2 adds a ladder of nested extracts of {{scale_ladder_sizes}} cells grown by one fixed rule. +Two committed extracts of MaleCNS v1.0 are used, streamed out of the ~1.1 GB public weight matrix on a CI runner and each under about 1 MB of JSON: **`named`**, a 1-hop neighbourhood of two seed cells ({{named_nodes}} nodes, {{named_edges}} edges), and **`taste_motor`**, the same seeds plus sensory types and an induced closure ({{taste_nodes}} nodes, {{taste_edges}} edges). A saved exploratory scale record describes nested extracts of {{scale_ladder_sizes}} cells; its larger input cuts are absent from this review package. -**The extracts are structurally very different, and the difference turns out to be load-bearing (T19).** `named` has mean degree {{named_mean_degree}}: {{named_share_onto_seeds}} of its edges terminate on the {{named_n_seed_nodes}} seed cells, only {{named_nodes_with_input}} of {{named_nodes}} nodes receive any input, and only {{named_share_recurrent_edges}} of edges leave a node that itself receives input. It is an in-star, not a circuit. `taste_motor` has mean degree {{taste_motor_mean_degree}}, {{taste_motor_share_onto_seeds}} of edges onto seeds and {{taste_motor_share_recurrent_edges}} of edges leaving a node with input. Composition differs too, and one class is absent from both ({{named_cells_octopamine}} cells), so a perturbation acting only through it cannot move either readout for substrate rather than domain reasons; the full census is in T19 and §S9(v). +**The extracts are structurally very different, and the difference turns out to be load-bearing (T19).** `named` has mean degree {{named_mean_degree}}: {{named_share_onto_seeds}} of its edges terminate on the {{named_n_seed_nodes}} seed cells, only {{named_nodes_with_input}} of {{named_nodes}} nodes receive any input, and only {{named_share_recurrent_edges}} of edges leave a node that itself receives input. It is dominated by edges entering seeds, with limited recurrent connectivity within this truncated cut. `taste_motor` has mean degree {{taste_motor_mean_degree}}, {{taste_motor_share_onto_seeds}} of edges onto seeds and {{taste_motor_share_recurrent_edges}} of edges leaving a node with input. Composition differs too, and one class is absent from both ({{named_cells_octopamine}} cells), so a perturbation acting only through it cannot move either readout for substrate rather than domain reasons; the full census is in T19 and §S9(v). Two runtimes consume a gain patch: a deterministic clipped leaky rate iteration (~50 ms per evaluation of `named`) and a current-based LIF network [7] whose synaptic scale and background drive are tuned constants, not fits (§S3). -**The rate runtime row-normalises its weight matrix, and previous versions of this paper did not say so.** Each row of the signed matrix is divided by `max(Σ|W|, 1)` before any gain is applied, and gains then act as a presynaptic column scale, so every node's recurrent input is a *composition-weighted average of its presynaptic gains*. That builds part of the answer to RQ2 into the readout itself — an example of the general hazard that a runtime convention can carry a result — so it is documented here, in T2 and §S3, and swept in §3.3 against `none` and a degree-corrected denominator. The sign table the runtime iterates is published as T21: {{sign_asserted}} of seven magnitudes ({{sign_asserted_transmitters}}) are asserted, varied by no member of the specification family and not bounded by the label-permuting null. +**The rate runtime row-normalises its weight matrix.** Each row of the signed matrix is divided by `max(Σ|W|, 1)` before any gain is applied, and gains then act as a presynaptic column scale. The resulting recurrent input depends on the weighted composition of a cell's presynaptic sources. This runtime convention can affect RQ2's classification; §3.3 compares it with no normalisation and a degree-corrected denominator. The sign table is reported in T21: {{sign_asserted}} of seven magnitudes ({{sign_asserted_transmitters}}) are asserted and fixed in the gain-rule family and Sobol analysis. The transmitter-label null changes label assignments, not these magnitudes. ### 2.3 Input-dependence testing (RQ2) @@ -96,93 +95,95 @@ The instrument we consider most transferable asks, for one `compound × concentr | `sign_permute_weight_matched` | + the real weights on the real edges, and each transmitter's share of total outgoing weight | | the MaleCNS cut | everything | -The rank-3 rung changed in this revision. Because a gain patch acts *through* transmitter identity, a null that moves the weighted balance moves the drug's target set and the sign matrix at once, so the plain label permutation is relabelled a **joint target-set-and-sign null** and replaced by a constrained shuffle holding every tracked transmitter's share of outgoing weight within {{bal_matched_max_deviation}} of the real graph's. Permuting the node transmitter-label array preserves the transmitter histogram *by count* but not *by weight*: out-strength is heterogeneous, so on the `named` cut the cholinergic share of total outgoing synaptic weight is {{bal_real_ach_share}} on the real graph against {{bal_plain_ach_mean}} ± {{bal_plain_ach_sd}} across {{bal_n}} permutations, putting the real graph at the {{bal_plain_ach_percentile}}th percentile of its own null (T23). +Because a gain patch acts *through* transmitter identity, the plain label-permutation null changes both the target set and the sign assignment. It is therefore analysed separately as a **joint target-set-and-sign null**. The rank-3 rung uses a constrained shuffle holding every tracked transmitter's share of outgoing weight within {{bal_matched_max_deviation}} of the real graph's. Permuting labels preserves the transmitter histogram *by count* but not *by weight*: out-strength is heterogeneous, so on the `named` cut the cholinergic share of total outgoing synaptic weight is {{bal_real_ach_share}} on the real graph against {{bal_plain_ach_mean}} ± {{bal_plain_ach_sd}} across {{bal_n}} permutations (T23). -**Non-rejection is not equivalence, and the software no longer lets the prose say it is.** Each mode receives one of three verdicts against a margin prespecified as {{dep_named_imidacloprid_delta_frac}} of the vehicle readout ({{dep_named_imidacloprid_delta}} Hz on `named`): `distinguishable` when the empirical two-sided permutation probability `(k+1)/(n+1)` is at or below α; `equivalent_within_tolerance` when the test does not reject *and* the gap from the null median is below the margin; `indeterminate` otherwise. From the verdicts follow a **class** and a **necessary information level** — the weakest graph model the test could not distinguish from the real cut, never a claim that it reproduces the effect. Where a weaker model survives while a richer one is rejected the ladder is non-monotone, which the software reports and §3.2 counts. +Each mode receives one of three verdicts against a margin prespecified as {{dep_named_imidacloprid_delta_frac}} of the vehicle readout ({{dep_named_imidacloprid_delta}} rate-model units on `named`): `distinguishable` when the empirical two-sided permutation probability `(k+1)/(n+1)` is at or below α; `equivalent_within_tolerance` when the test does not reject *and* the observed point gap from the null median is below the margin; `indeterminate` otherwise. The point-gap verdict is descriptive, not a formal equivalence test with a confidence bound. A **class** and **necessary information level** summarise these component verdicts. The level identifies the weakest graph model this test could not distinguish from the real cut; it does not establish that the model reproduces the effect. Non-monotone ladders are flagged and counted in §3.2. -Three further conventions are prespecified and §S2 states them in full: a *relative* effect floor, a separation of **confirmatory** from **exploratory** tests with Benjamini-Hochberg applied across the latter, and a check of how many tests sit at the resolution floor, so a corrected count of zero is never mistaken for a finding. +Three further conventions are specified in §S2: a *relative* effect floor, separate planned profiles and exploratory landscapes, and a check of permutation resolution. Benjamini-Hochberg adjusts the exploratory component-test probabilities within the declared family; it does not provide a false-discovery guarantee for the derived cell classes. ### 2.4 The ablation ladder and the specification family (RQ3) *Which layer carries the information?* Four models predict the same cell with one layer denied: **A** receptor engagement alone; **B** mechanism gains on the graph's transmitter composition only, no edges; **C** the real cut and runtime with one generic multiplier; **D** the full model. They carry different units, so they are compared by how well each reproduces the full model's *ordering* of the library, scored on the **signed** Spearman ρ because a perfect inversion reproduces nothing. -Level C's multiplier changed in this revision: it could previously only *depress*, so no entry could come out positive and the level could not order a library half of whose entries raise the readout. Its magnitude is unchanged and its **sign** now comes from the mechanism table — one bit per compound — with the old rule retained as a floor (§S1). Levels B and D are not independent models, both being functions of the same gain vector, so §3.3 reports a **matched reference distribution** saying what agreement would have been for structural reasons alone. +Level C uses a generic magnitude with a mechanism-derived sign, one bit per compound. A depression-only generic rule is retained as a sensitivity reference (§S1). Levels B and D are not independent models, both being functions of the same gain vector, so §3.3 reports a **matched reference distribution** saying what agreement would have been for structural reasons alone. -*Which conclusions are properties of the data rather than of a modelling choice?* The parameter-to-gain mapping is the least defensible part of the pipeline — a textbook shape, but an asserted functional form. We prespecify {{stab_family_size}} admissible specifications (five shapes at five coefficient scales, §S1.1) and re-derive every principal conclusion under every member; undecidable cells count against the conclusion and are reported separately, and the fraction is a fraction of *that family*, not a posterior probability. The topology conclusions are decided at {{stab_n_shuffles}} permutations per specification (resolution {{stab_shuffle_resolution}}), corrected across the run's {{stab_n_structural_tests}} structural tests. The two thresholds are likewise conventions, so the verdict is recomputed over the full {{thr_circuit_fracs}} × {{thr_vert_limits}} grid on {{thr_n_compounds}} compounds. +*Which model-output predicates depend on the gain rule?* The parameter-to-gain mapping is the least defensible part of the pipeline — a textbook shape, but an asserted functional form. We evaluate {{stab_family_size}} gain-rule specifications (five shapes at five coefficient scales, §S1.1) and re-derive each stated model-output predicate under every member; undecidable cells count against retention and are reported separately, and the fraction is a fraction of *that family*, not a posterior probability. The topology predicates are decided at {{stab_n_shuffles}} permutations per specification (resolution {{stab_shuffle_resolution}}), corrected across the run's {{stab_n_structural_tests}} structural tests. The two thresholds are likewise conventions, so the verdict is recomputed over the full {{thr_circuit_fracs}} × {{thr_vert_limits}} grid on {{thr_n_compounds}} compounds. ### 2.5 Evaluation taxonomy -Agreement with published work is worth little unless its kind is named, so we separate three. **Implementation consistency** — the same inputs give the same outputs across transports, runtimes and runs (§3.7). **Literature concordance** — the model's *ordering* agrees with a published ordering, scored per entry and never as a bare mean, since an entry ordering two items can only return ±1; where the published ordering shares a source with the library, agreement is internal consistency, and the pipeline computes that overlap from DOIs and PMIDs rather than asserting it. **Independent out-of-sample validation** — against data entering neither the library nor the model, of which there is **none** at the circuit level. The ground-truth experiment of §3.1 validates the *instrument*, not the domain model, and we are careful not to let the first stand in for the second. +Agreement with published work is worth little unless its kind is named, so we separate three. **Implementation consistency** — the same inputs give the same outputs across transports, runtimes and runs (§3.7). **Literature concordance** — the model's *ordering* agrees with a published ordering, scored per entry and never as a bare mean, since an entry ordering two items can only return ±1; where the published ordering shares a source with the library, agreement is internal consistency, and the pipeline computes that overlap from DOIs and PMIDs rather than asserting it. **Independent out-of-sample validation** — against data entering neither the library nor the model, of which there is **none** at the circuit level. The planted-cycle experiment of §3.1 checks recovery for one synthetic mechanism, not the domain model. ### 2.6 Global uncertainty and the value of information -A one-at-a-time tornado cannot separate a factor's own effect from its interactions, so uncertainty is attributed with a variance-based (Sobol') analysis over every assumption the implementation exposes — potency, Hill coefficient, the engagement-to-gain transformation and its coefficient, the external drive, the cut's synapse-weight threshold, the transmitter-sign convention, the expression-weighting layer — plus a deliberately null factor, the spiking RNG seed, which the deterministic rate engine ignores. The design is Saltelli cross-sampling with Jansen estimators [71], [72], validated against the analytic Ishigami benchmark [73] in the same run. +A one-at-a-time sensitivity analysis cannot separate a factor's own effect from interactions, so uncertainty is attributed with a variance-based (Sobol') analysis over the selected factors: potency, Hill coefficient, engagement-to-gain transformation and coefficient, external drive, the cut's synapse-weight threshold, transmitter-label reassignment probability, expression weighting and a deliberately null spiking RNG seed, which the deterministic rate engine ignores. Transmitter-label reassignment does not vary the fixed sign magnitudes in T21. The design uses Saltelli cross-sampling with Jansen estimators [70], [71] and is checked against the analytic Ishigami benchmark [72]. -Estimates are left unclipped, so a factor whose true index is zero can be estimated below zero. A first-order index cannot be negative, so the **largest magnitude among the negative estimates is a measured lower bound on the estimator's error**, and that — not the null factor's own draw — is the noise floor we report. The budget becomes a ranking by `VOI_j = max(0, S_j) · Var(Y)` [76] — variance of a model output under assumed input ranges, not an expected gain in accuracy about a living fly. +Estimates are left unclipped, so a factor whose true index is zero can be estimated below zero. A first-order index cannot be negative, so the **largest magnitude among the negative estimates is a measured lower bound on the estimator's error**, and that — not the null factor's own draw — is the noise floor we report. For prioritisation we define a heuristic score `H_j = max(0, S_j) · Var(Y)`, conceptually motivated by information-value reasoning [75]. It ranks model-output variance under assumed ranges, not formal expected information value or biological accuracy. ## 3. Results -This section evaluates the instruments, not the domain. All values are regenerated by `scripts/reproduce_paper.py` and stored keyed in `papers/results.json`; the parameter library and the domain readouts they are demonstrated on are described in §3.6 and §S6. +This section evaluates the instruments. Core analysis values are regenerated by `scripts/reproduce_paper.py` and stored keyed in `papers/results.json`; larger scale-study values are imported from a saved record. The parameter library and the domain readouts they are demonstrated on are described in §3.6 and §S6. -### 3.1 Is the dependence instrument sound? Ground truth, power and convergence (Fig. F16, Tables T11, T22) +### 3.1 Synthetic positive controls and permutation convergence (Fig. F16, Tables T11, T22) -A negative from a dependence test is uninterpretable unless the test can be shown to return a positive when one exists, so the instrument is characterised against **planted ground truth** before it is used on anything. A recurrent cycle of known strength is planted in a synthetic network of the same size, density and composition as the real extract, so that the perturbation removes an amplification existing only while the cycle is intact: the *contrast*, not merely the level, depends on the wiring by construction. Strength 0 plants nothing and is the negative control. +The dependence procedure is checked against a **planted synthetic contrast**. A recurrent cycle of known strength is planted in a synthetic network of {{val_recovery_nodes}} nodes and about {{val_recovery_edges}} background edges. Its transmitter composition was chosen close to the `named` cut, but size and connectivity were not matched. The perturbation removes amplification that exists while the planted cycle is intact, so the treatment contrast depends on its wiring by construction. Strength 0 plants nothing and is the negative control. This design tests recovery for the planted mechanism, not power for all network effects. -The ladder recovered {{val_recovery_detected}} of {{val_recovery_positive}} planted effects at *n* = {{val_recovery_n}} with no false positive on the control. Over a power grid ({{val_power_replicates}} independently generated networks per cell, permutation counts {{val_power_ns}}) the detection rate by planted strength is {{val_power_by_strength}} — full power at strength {{val_power_full_strengths}}, and at or below one in four at {{val_power_weak_strengths}}, so the instrument is **blind to weak planted structure and must not be read as ruling it out**. The empirical false-positive rate on the unplanted control is {{val_power_control_hits}} of {{val_power_control_runs}}. Averaged over strengths, detection by permutation count is {{val_power_by_n}}: **effect size dominates, and permutation count buys little beyond a few hundred draws.** That is a design rule for this class of test — budget spent on permutations past the point of convergence is wasted, and it is the reason the scaling study of §3.2 can trade permutations for network size at all. +The ladder recovered {{val_recovery_detected}} of {{val_recovery_positive}} planted effects at *n* = {{val_recovery_n}} with no false positive on the control. In a small grid of {{val_power_replicates}} independently generated networks per cell and permutation counts {{val_power_ns}}, detection by planted strength was {{val_power_by_strength}}; the empirical false-positive count was {{val_power_control_hits}} of {{val_power_control_runs}}. Detection was limited at {{val_power_weak_strengths}}. The count was {{val_power_by_n}} across permutation budgets. The observed full-detection strengths are {{val_power_full_strengths}} in this finite grid. These synthetic results motivate a sensitivity warning for weak effects; they do not establish general power or justify a universal permutation budget. The convergence sweep (T11) separates the two quantities that permutation budget buys: a *verdict* at α = 0.05 settles from *n* = {{dep_verdict_stable_from_n}}–{{dep_verdict_stable_to_n}}, while a quotable mid-range probability needs {{dep_p_within_0p02_from_n}}–{{dep_p_within_0p02_worst_n}} draws to reach ±0.02 of its final value. The sweep re-reads prefixes of the same draws, so every row is a valid smaller sample and no extra runtime was spent producing it. -### 3.2 The verdict inverts with the extract, and recurrence predicts it (Figs. F6, F12; Tables T6, T12, T12b, T23, T27) +### 3.2 Dependence labels differ between the two extracts (Figs. F6, F12; Tables T6, T12, T12b, T23, T27) -**The confirmatory profiles** (full per-mode table: T6, §S2.1). At {{dep_n}} permutations per mode (resolution {{dep_p_floor}}), imidacloprid's effect on the sparse `named` extract is {{dep_named_imidacloprid_effect}} Hz and the cell is **{{dep_named_imidacloprid_class}}**: against the prespecified margin of {{dep_named_imidacloprid_delta}} Hz it is equivalent within tolerance for {{dep_named_imidacloprid_equivalent_modes}}, merely indeterminate for {{dep_named_imidacloprid_indeterminate_modes}}, and distinguishable only from the Erdős–Rényi control (*p* = {{dep_named_imidacloprid_p_erdos_renyi}} against {{dep_named_imidacloprid_p_rewire_degree_preserving}} for degree-preserving rewiring). Fipronil on the same extract is the contrast — {{dep_named_fipronil_effect}} Hz, **{{dep_named_fipronil_class}}**, *p* = {{dep_named_fipronil_p_weight_permute}} and {{dep_named_fipronil_p_rewire_degree_preserving}} for weight permutation and rewiring, necessary level `{{dep_named_fipronil_level}}` — so the instrument separates two predictions on one extract. Correcting the transmitter null moves no probability but changes what may be claimed: imidacloprid's gap from the null median falls from {{dep_named_imidacloprid_gap_sign_permute}} Hz to {{dep_named_imidacloprid_gap_sign_permute_weight_matched}} Hz, turning an indeterminate result into an equivalence one. +**The planned profiles** (full per-mode table: T6, §S2.1). At {{dep_n}} permutations per mode (resolution {{dep_p_floor}}), imidacloprid's effect on the sparse `named` extract is {{dep_named_imidacloprid_effect}} rate-model units and the cell is **{{dep_named_imidacloprid_class}}**: against the declared margin of {{dep_named_imidacloprid_delta}} rate-model units, {{dep_named_imidacloprid_equivalent_modes}} null comparisons pass the descriptive point-gap rule, {{dep_named_imidacloprid_indeterminate_modes}} remain unresolved, and the Erdős–Rényi control is distinguishable (*p* = {{dep_named_imidacloprid_p_erdos_renyi}} versus {{dep_named_imidacloprid_p_rewire_degree_preserving}} for degree-preserving rewiring). Fipronil on the same extract is the contrast — {{dep_named_fipronil_effect}} rate-model units, **{{dep_named_fipronil_class}}**, *p* = {{dep_named_fipronil_p_weight_permute}} and {{dep_named_fipronil_p_rewire_degree_preserving}} for weight permutation and rewiring, reported level `{{dep_named_fipronil_level}}`. The weight-matched transmitter null reduces imidacloprid's gap from {{dep_named_imidacloprid_gap_sign_permute}} to {{dep_named_imidacloprid_gap_sign_permute_weight_matched}} rate-model units, changing an unresolved comparison to a point-gap pass without establishing statistical equivalence. -**The landscape inverts between the two extracts.** Over {{dep_land_compounds}} compounds × {{dep_land_concs}} concentrations = {{dep_land_cells}} cells at {{dep_n_landscape}} permutations, with Benjamini-Hochberg across the {{dep_land_n_structural_tests}} structural tests and the relative effect floor applied, `named` gives **{{dep_land_topology_dependent}} topology-dependent, {{dep_land_composition_dominated}} composition-dominated, {{dep_land_no_effect}} no-effect, {{dep_land_mixed}} mixed** (uncorrected {{dep_land_topology_dependent_raw}} / {{dep_land_composition_dominated_raw}} / {{dep_land_no_effect_raw}}; on the absolute floor alone {{dep_land_topology_dependent_absfloor}} / {{dep_land_composition_dominated_absfloor}} / {{dep_land_no_effect_absfloor}}, the {{dep_land_below_relative_floor}} cells the relative floor removes named in T12). Repeated unchanged on the denser `taste_motor` extract at {{dep_n_landscape_taste}} permutations, the same {{dep_land_taste_cells}} cells give **{{dep_land_taste_topology_dependent}} / {{dep_land_taste_composition_dominated}} / {{dep_land_taste_no_effect}}**: every cell above the floor is topology-dependent, imidacloprid included (*p* = {{dep_land_taste_imidacloprid_p_rewire_degree_preserving}}, adjusted {{dep_land_taste_imidacloprid_q}}, class {{dep_land_taste_imidacloprid_class}}). +**Descriptive landscape.** The `named` cut has {{named_share_onto_seeds}} of edges entering four seed cells. Over {{dep_land_compounds}} compounds × {{dep_land_concs}} concentrations = {{dep_land_cells}} cells at {{dep_n_landscape}} permutations, with Benjamini-Hochberg across the {{dep_land_n_structural_tests}} structural component tests and the relative effect floor applied, it gives **{{dep_land_topology_dependent}} topology-dependent, {{dep_land_composition_dominated}} composition-dominated, {{dep_land_no_effect}} no-effect, {{dep_land_mixed}} mixed** labels (uncorrected {{dep_land_topology_dependent_raw}} / {{dep_land_composition_dominated_raw}} / {{dep_land_no_effect_raw}}; on the absolute floor alone {{dep_land_topology_dependent_absfloor}} / {{dep_land_composition_dominated_absfloor}} / {{dep_land_no_effect_absfloor}}, with {{dep_land_below_relative_floor}} cells removed by the relative floor in T12). The denser `taste_motor` extract uses {{dep_n_landscape_taste}} permutations and gives **{{dep_land_taste_topology_dependent}} / {{dep_land_taste_composition_dominated}} / {{dep_land_taste_no_effect}}** among the same {{dep_land_taste_cells}} cells; these classifications have no cell-level FDR guarantee and are conditional on each cut's nulls and runtime. -**Recurrence, not size, predicts the inversion.** Repeating the same profile across a ladder of nested extracts grown from one seed set by a single fixed rule ({{scale_study_n_cuts}} extracts in all, {{scale_study_cuts}}; T27) settles the question of what the reversal tracks. The composition-dominated verdict appears on exactly {{scale_study_n_composition_dominated}} of them, `{{scale_study_composition_dominated_cuts}}`, and {{scale_study_density_not_size}}. Imidacloprid is topology-dependent on every rung but the in-star, and fipronil on all seven (per-rung sequences in §S2.4, full table T27). Both compounds' verdicts are settled on the clean ladder by the two-largest-rungs criterion of §S2.4, and at the denser middle rungs ({{scale_study_every_null_cuts}}) the prediction is distinguishable from **every** null, its necessary level being the real network itself. The relative effect size is stable over this range ({{scale_study_imidacloprid_relative_range}} of the vehicle readout), so the change of verdict is not a change of what is being measured. +**Exploratory scale record.** The saved scale study reports {{scale_study_n_cuts}} cuts ({{scale_study_cuts}}; T27). The composition-dominated label appears on {{scale_study_n_composition_dominated}} ({{scale_study_composition_dominated_cuts}}), with per-rung sequences in §S2.4. The larger 5k–50k input cuts are not in this package, so those rungs cannot be independently rebuilt from the supplied data. The observed association of cut construction, density and label does not isolate recurrence as a cause; the two committed cuts also differ in composition and extraction rule. The lower permutation budgets on larger cuts ({{scale_study_n_range}} draws) further limit comparisons. -Two honesty constraints travel with that result. The permutation budget must fall as the network grows ({{scale_study_n_range}} draws, resolution {{scale_study_resolution_range}}), and {{scale_study_coarse_note}} — so the top rungs confirm the settled verdict rather than establishing it. And the two hops-limited cuts are built by a different rule from the nested rungs, so only comparisons *within* the ladder separate scale from construction. +The scale record uses decreasing permutation budgets as edge count grows ({{scale_study_n_range}} draws; resolution {{scale_study_resolution_range}}). Comparisons between the hops-limited committed cuts and nested ranked-BFS scale cuts confound size with extraction rule. Results from the missing larger inputs are reported as imported observations, not reproduced outcomes. -**What is withdrawn, and what replaces it.** A degree-preserving rewire of an in-star is close to the identity, so the null had nothing to destroy; an extract with genuine recurrence does, and destroying it changes the answer. We withdraw the generalisation that most of what such a model predicts is carried by composition rather than topology, and the domain claim that went with it. What replaces it is stronger and methodological: **a dependence verdict is a joint property of the prediction and the extract, the extract's recurrence predicts which way it goes, and a negative measured on a sparse extract is not evidence about the network it was cut from.** +**Interpretation.** The sparse `named` cut has a large incoming-star component, whereas `taste_motor` contains more edges among nodes receiving input (T19). Degree-preserving rewiring may therefore change less of the effective input pattern on `named`. This is one plausible explanation for the contrasting labels, not a causal test of recurrence. Dependence classifications must be reported with the extract and runtime used. -**Multiplicity, working.** As a prespecified confirmatory profile, fipronil at 1 µM stands comfortably (*p* = {{dep_named_fipronil_min_structural_p}} at *n* = {{dep_n}}). Inside the exploratory landscape the same cell's adjusted probability is {{dep_land_fipronil_q}} and it does **not** survive correction; its corrected class there is {{dep_land_fipronil_class}} against an uncorrected {{dep_land_fipronil_class_raw}}. +**Multiplicity.** The planned fipronil profile at 1 µM has component-test *p* = {{dep_named_fipronil_min_structural_p}} at *n* = {{dep_n}}. In the exploratory landscape, a corresponding adjusted component-test value is {{dep_land_fipronil_q}}, with corrected class {{dep_land_fipronil_class}} versus {{dep_land_fipronil_class_raw}} before adjustment. Benjamini-Hochberg applies to the declared structural test family; the cell class itself has no separate FDR guarantee. -**Two things the class label does not say.** The ladder is **non-monotone** in {{dep_land_non_monotone}} of {{dep_land_cells}} cells on `named` and {{dep_land_taste_non_monotone}} of {{dep_land_taste_cells}} on `taste_motor`, because adjacent rungs destroy different things; the level therefore means "the cheapest degraded input this test cannot tell apart from the real one", never "everything above it is indistinguishable too". And `composition-dominated` reports a failure to distinguish, never a reproduction: across the `named` landscape's mode verdicts, {{dep_land_verdicts_equivalent_within_tolerance}} are equivalent within tolerance, {{dep_land_verdicts_indeterminate}} indeterminate and {{dep_land_verdicts_distinguishable}} distinguishable. +The planned fipronil profile and its exploratory landscape class therefore answer different testing questions; a nominal result in the profile does not override the adjusted landscape label. + +**What the class label does not say.** The ladder is **non-monotone** in {{dep_land_non_monotone}} of {{dep_land_cells}} cells on `named` and {{dep_land_taste_non_monotone}} of {{dep_land_taste_cells}} on `taste_motor`, because adjacent rungs destroy different things. The level is the lowest-ranked degradation that passes the declared test and point-gap rule, not a monotone information threshold. A `composition-dominated` label also does not mean the null reproduces the effect: across `named` mode verdicts, {{dep_land_verdicts_equivalent_within_tolerance}} pass the descriptive point-gap rule, {{dep_land_verdicts_indeterminate}} are unresolved and {{dep_land_verdicts_distinguishable}} are distinguishable. ### 3.3 Which pipeline layer carries the information, and what agreement would have been anyway (Fig. F13, Tables T13, T13b, T24, T25) -Receptor engagement alone orders the library **backwards** (signed ρ {{abl_rho_receptor_min}} to {{abl_rho_receptor_max}}) and is now the worst level at every concentration — constant across the table: {{abl_worst_level_constant}} ({{abl_worst_level_by_conc}}). Under the historical level-C rule the worst level was *not* constant ({{abl_worst_level_by_conc_floor_rule}}), so the previous claim that the topology-only level is the worst of the four was true of one column and false of the others. +Under the current library, receptor engagement alone has signed rank correlation ρ {{abl_rho_receptor_min}} to {{abl_rho_receptor_max}} with the full-model ordering. The lowest-correlation level is {{abl_worst_level_constant}} across the reported concentrations ({{abl_worst_level_by_conc}}); under the depression-only level-C reference, the lowest level varies ({{abl_worst_level_by_conc_floor_rule}}). This comparison is conditional on the library and generic rule. -**The ablation conclusion is withdrawn.** Level C's multiplier applied a depression-only factor to every transmitter, so every entry was at most zero and the level could not represent disinhibition. With a direction-aware rule — same generic magnitude, one bit of sign — its rank correlation with the full model at 1 µM rises from {{abl_rho_topology_floor_paper}} to {{abl_rho_topology_paper}}. The sentence "the connectome without the pharmacology is not a cheap substitute for the connectome with it" was a statement about a sign-broken baseline and is retracted. +**Ablation reference.** A depression-only generic multiplier cannot represent disinhibition. The direction-aware rule uses the same generic magnitude with a mechanism-derived sign, and its topology-only rank correlation at 1 µM is {{abl_rho_topology_paper}} versus {{abl_rho_topology_floor_paper}} under the depression-only reference. The difference shows that the generic rule affects the rank comparison. -**The composition-versus-full correlation carries no compound-level information.** It is {{abl_rho_composition_paper}} at 1 µM and {{abl_ref_no_floor}} after dropping the compounds on the 0.05 gain floor, so it is not a tie artefact. But levels B and D are both functions of the same gain vector, and the matched reference says what would have been unsurprising: pharmacology-free pseudo-compounds with the gain shape the shipped rules produce give a median of {{abl_ref_matched_median}} over {{abl_ref_n_draws}} draws (5th–95th percentile {{abl_ref_matched_p05}} to {{abl_ref_matched_p95}}), and permuting the compound labels leaves the observed value identically {{abl_ref_shuffled}}, as it must, because the permutation moves the (B, D) pairs as a block. +**The composition-versus-full correlation does not by itself establish compound-level information.** It is {{abl_rho_composition_paper}} at 1 µM and {{abl_ref_no_floor}} after dropping compounds on the 0.05 gain floor, so it is not solely a tie artefact. Levels B and D both use the same gain vector. Pharmacology-free pseudo-compounds matched to the shipped gain shapes give a reference median of {{abl_ref_matched_median}} over {{abl_ref_n_draws}} draws (5th–95th percentile {{abl_ref_matched_p05}} to {{abl_ref_matched_p95}}). Permuting compound labels moves B–D pairs together and leaves the observed correlation {{abl_ref_shuffled}}. -**And it is not robust to the runtime.** The correlation is {{abl_rho_composition_row_abs}} under the shipped normalisation and {{abl_rho_composition_none}} with none, but {{abl_rho_composition_degree}} under a degree-corrected denominator at 1 µM and {{abl_rho_composition_degree_1e8}} at 10 nM: it survives {{abl_composition_survives_normalisations}} and not a degree correction, so it is a property of *this runtime's readout* as much as of the perturbation. A sign disagreement between the two levels moves it by at most {{abl_glutamate_max_delta_rho}} and so is not what produces the agreement (§S9(i)). +**Runtime normalisation changes the agreement.** The correlation is {{abl_rho_composition_row_abs}} under the shipped normalisation and {{abl_rho_composition_none}} with none, but {{abl_rho_composition_degree}} under a degree-corrected denominator at 1 µM and {{abl_rho_composition_degree_1e8}} at 10 nM. Agreement holds under row normalisation and without normalisation but weakens under degree correction, so it depends on the runtime's normalisation. A sign disagreement between the two levels moves it by at most {{abl_glutamate_max_delta_rho}} and so is not what produces the agreement (§S9(i)). -### 3.4 Specification robustness: which conclusions survive the modelling choices (Fig. F14, Tables T14, T14b, T15) +### 3.4 Gain-rule sensitivity of model-output predicates (Fig. F14, Tables T14, T14b, T15) -**The previous 25/25 topology result was vacuous, and the failure mode is general.** Two faults compounded. The matrix was computed at 6 shuffles per specification, where the smallest attainable permutation probability is 1/7 = 0.143, so `p ≤ 0.05` was arithmetically unattainable and the predicate silently fell back on a Gaussian criterion. And the specification context rebound the transformation on the assay modules but not on the engine the permutation path resolves, so **all {{stab_family_size}} specifications fed default values to their nulls and produced identical effects**. A robustness matrix can thus report perfect stability because its alternatives never reached the code under test — a bug no determinism check can catch, since the run is perfectly reproducible either way. +**Specification test conditions.** The topology predicates use {{stab_n_shuffles}} shuffles per specification (resolution {{stab_shuffle_resolution}}), and the gain specification is passed to the null engine. These conditions allow the empirical test to reach α = 0.05 and ensure the intended mapping is evaluated. The reported retention fractions are conditional on this implementation and the declared 25-member family. -Recomputed at {{stab_n_shuffles}} shuffles per specification (resolution {{stab_shuffle_resolution}}), with the specification pushed into the null engine and Benjamini-Hochberg applied across the {{stab_n_structural_tests}} structural tests, both topology conclusions still hold {{stab_C3_retained}}/{{stab_family_size}} and {{stab_C4_retained}}/{{stab_family_size}}. The nuance the old matrix could not show is that a retention rate for a conclusion shaped as a *failure to reject* is only as strong as the equivalence behind it: of C3's {{stab_C3_retained}}, {{stab_C3_equivalent}} reach equivalence within the prespecified margin and {{stab_C3_indeterminate}} are merely indeterminate. Under the {{stab_family_size}} specifications: +Across {{stab_family_size}} gain-rule specifications, C3 and C4 are retained in {{stab_C3_retained}} and {{stab_C4_retained}} cases. These C1–C7 conclusions are predicates over model outputs and the chosen null procedure, not robust biological conclusions. The transmitter sign magnitudes and transmitter-defined assignment are fixed across the family. For C3, {{stab_C3_equivalent}} of the retained cases meet the descriptive point-gap criterion and {{stab_C3_indeterminate}} are indeterminate. Benjamini-Hochberg adjusts the {{stab_n_structural_tests}} component tests in the run; it does not guarantee a false-discovery rate for predicate retention. -* The domain's most quotable claim — that the agonist suppresses activity — is retained by only {{stab_C1_retained}} and **reversed** by all {{stab_C1_lost}} monotone specifications: the same compound at the same engagement takes the network from {{stab_monotone_vehicle_hz}} to {{stab_monotone_treated_hz}} Hz ({{stab_monotone_pct_change}}) instead of from {{stab_default_vehicle_hz}} to {{stab_default_treated_hz}} Hz, with `g_ach` at {{stab_monotone_g_ach}} against {{stab_default_g_ach}}. The direction is a property of one asserted term, not of the integration. -* A buffering claim is retained {{stab_C5_retained}} of {{stab_family_size}} and reversed by none, its {{stab_C5_undecidable}} failures being undecidable. The previous explanation — that the monotone rules buffer "at least as much, never less" — is **false**: over the {{stab_mean_gap_nicotinic_monotone_n}} reversing specifications with a defined index the mean gap runs {{stab_mean_gap_nicotinic_monotone_max}} to {{stab_mean_gap_nicotinic_monotone_min}} log₁₀ units, {{stab_monotone_buffer_less_n}} buffer *less* ({{stab_monotone_buffer_less}}), and the aggregate is carried by the extreme — an aggregate standing in for a range it does not represent. +* The domain's most quotable claim — that the agonist suppresses activity — is retained by only {{stab_C1_retained}} and **reversed** by all {{stab_C1_lost}} monotone specifications: the same compound at the same engagement takes the network from {{stab_monotone_vehicle_hz}} to {{stab_monotone_treated_hz}} rate-model units ({{stab_monotone_pct_change}}) instead of from {{stab_default_vehicle_hz}} to {{stab_default_treated_hz}} rate-model units, with `g_ach` at {{stab_monotone_g_ach}} against {{stab_default_g_ach}}. The direction is a property of one asserted term, not of the integration. +* The buffering predicate is retained in {{stab_C5_retained}} of {{stab_family_size}} cases and undecidable in {{stab_C5_undecidable}}. Across {{stab_mean_gap_nicotinic_monotone_n}} monotone specifications with a defined index, the mean gap spans {{stab_mean_gap_nicotinic_monotone_max}} to {{stab_mean_gap_nicotinic_monotone_min}} log₁₀ units; {{stab_monotone_buffer_less_n}} buffer less than the shipped rule ({{stab_monotone_buffer_less}}). * The matching amplification claim is weakest at {{stab_C6_retained}} of {{stab_family_size}}. Retained by every specification: {{stab_specification_independent}}. The threshold grid makes the same point about conventions rather than transformations: the split is **not** stable (`stable = {{thr_stable}}`), buffering holding in {{thr_nicotinic_matching}} of {{thr_nicotinic_cells}} cells ({{thr_nicotinic_gap_max}} to {{thr_nicotinic_gap_min}} log₁₀ units) and amplification in {{thr_nav_ache_matching}} of {{thr_nav_ache_cells}}, flipping at the strictest corner ({{thr_nav_ache_failing}}, gap {{thr_nav_ache_gap_min}}), so ranges are quoted and never the point estimate. Two tables that appear to disagree in fact score different sets — T15 the {{thr_n_compounds}} compounds of two mechanism classes, T8 the whole library — so {{thr_outside_t15_scope}} is scored in one and outside the other's scope. ### 3.5 Variance attribution, and what to measure next (Fig. F15, Tables T16, T17) -The Sobol' analysis ({{unc_evaluations}} model evaluations at n_base = {{unc_n_base}}, Var(Y) = {{unc_variance}} Hz²) resolves {{unc_n_resolved}} of {{unc_n_factors}} factors in the strict sense of a bootstrap first-order interval excluding zero: {{unc_resolved}} — the **engagement-to-gain transformation** (S1 {{unc_S_gain_transform}}, ST {{unc_ST_gain_transform}}) and the cut's **synapse-weight threshold** (S1 {{unc_S_weight_threshold}}, ST {{unc_ST_weight_threshold}}). Everything else is unresolved at this sample size ({{unc_at_noise_floor}}), including the external drive, whose S1 of {{unc_S_drive}} sits below the estimator's noise floor of {{unc_noise_floor}} on an interval of {{unc_CI_drive}}. The floor is set by {{unc_noise_floor_factor}}, not by the declared null factor `lif_seed` at {{unc_null_factor_S}}; the previous version of this section quoted the null factor and then ranked a factor inside the resulting noise. Interactions carry {{unc_interaction_share}} of the variance raw and {{unc_interaction_share_clipped}} with negatives clipped. On the analytic benchmark the estimator returns {{ish_estimated}} against the exact {{ish_analytic}} at n_base = {{ish_n_base}} (max error {{ish_max_abs_error}}). +The Sobol' analysis uses {{unc_evaluations}} model evaluations at n_base = {{unc_n_base}} and yields Var(Y) = {{unc_variance}} in squared rate-model units. Bootstrap first-order intervals exclude zero for {{unc_n_resolved}} of {{unc_n_factors}} factors ({{unc_resolved}}), including the engagement-to-gain transformation (S1 {{unc_S_gain_transform}}, ST {{unc_ST_gain_transform}}) and cut weight threshold (S1 {{unc_S_weight_threshold}}, ST {{unc_ST_weight_threshold}}). Other factors are unresolved at this sample size ({{unc_at_noise_floor}}). The first-order estimate for external drive is {{unc_S_drive}} with interval {{unc_CI_drive}}, and the negative-estimate noise diagnostic is {{unc_noise_floor}}, driven by {{unc_noise_floor_factor}} rather than the deliberately null factor (estimate {{unc_null_factor_S}}). Estimated interaction shares are {{unc_interaction_share}} raw and {{unc_interaction_share_clipped}} after clipping negative estimates. Ishigami estimates are {{ish_estimated}} versus analytic values {{ish_analytic}} (maximum absolute error {{ish_max_abs_error}}). These numerical diagnostics do not validate the biological model. -The top factor is **{{voi_rank1_factor}}** ({{voi_rank1_var}} of {{voi_var_total}} Hz²), resolved by *{{voi_rank1_experiment}}* — reading the *shape* of the response-versus-engagement curve at a known synapse, which is precisely the assumption §3.4 shows the suppression conclusion turns on. The second, **{{voi_rank2_factor}}** ({{voi_rank2_var}} Hz²), needs **no experiment at all**: it is resolved by re-analysing synapse-confidence strata of data already held. +The author-defined variance-priority heuristic ranks **{{voi_rank1_factor}}** first ({{voi_rank1_var}} of {{voi_var_total}} squared rate-model units). A direct assay of the response-versus-engagement curve at a known synapse ({{voi_rank1_experiment}}) could constrain this factor. **{{voi_rank2_factor}}** ranks second ({{voi_rank2_var}} squared rate-model units) and could be constrained by re-analysis of synapse-confidence strata. These rankings are conditional on assumed factor ranges, not formal expected values of information or evidence of biological importance. ### 3.6 What the refusal rule actually does (Fig. F11, Tables T10, T20) -The type system changes results rather than documentation. Of {{ev_rows}} library rows, {{ev_rows_not_modelled}} carry no admissible evidence and return N/A instead of a small plausible number, and the admissible transformation is decided by the (parameter type, distance) pair rather than by the author: {{ev_model_binding_occupancy}} row earns a physical binding model, {{ev_model_binding_engagement_proxy}} a labelled binding proxy, {{ev_model_functional_engagement}} a direct functional engagement and {{ev_model_functional_engagement_proxy}} a transferred one. The single physical row is {{ev_binding_row_names}} [77]; the library's previous binding constant, {{ev_binding_proxy_row_names}}, is demoted to a cross-species proxy — the same number under a name that says it was transferred. Removing that one row would leave a branch of the type system with no instances, which is the kind of fact a lookup table hides and an exhaustively checked invariant surfaces. The full census, the domain readouts the engagements feed, and the literature-concordance analysis are in §S6 and §S10. +The type system changes which rows enter numerical predictions. Of {{ev_rows}} library rows, {{ev_rows_not_modelled}} are withheld and return N/A. The admitted categories (T20) contain physical binding ({{ev_model_binding_occupancy}}), binding proxy ({{ev_model_binding_engagement_proxy}}), direct functional engagement ({{ev_model_functional_engagement}}) and transferred functional engagement ({{ev_model_functional_engagement_proxy}}) rows, assigned by declared type and distance. The binding model row is {{ev_binding_row_names}} [76]; {{ev_binding_proxy_row_names}} is a transferred proxy. These admission categories do not independently verify every numeric source-value link. A row-level audit found assay mismatches and demoted affected entries; remaining numerical entries require source-to-parameter verification (§S6). ### 3.7 Reproduction, determinism and cross-backend parity (Fig. F1) -One command regenerates every figure, table and number in this paper from committed artifacts, with no download and no access to the 1.1 GB weight matrix: +The reproduction command regenerates the main analysis and its figures and tables from committed artifacts without downloading the 1.1 GB weight matrix. The saved scale-study record includes larger input cuts absent from the current package: ```bash python -m pip install -e ".[dev,viz]" @@ -191,199 +192,203 @@ python scripts/reproduce_paper.py --fast # reduced permutation and replicate c python scripts/reproduce_paper.py --only dependence --outdir /tmp/x ``` -Beyond the paper, `flylab run` turns a declarative spec into a self-describing run directory whose manifest records the spec, code and library hashes, the seeds and two hashes per artifact; the same spec and seed reproduce it byte for byte (`docs/WORKFLOW.md`). `papers/results.json` carries every scalar quoted above with its value, rendered text, unit and producing step; two runs at one commit and seed produce identical values, runtimes excepted, and a test asserts it. Every numeric span in the manuscript is a substitution from that record, so a number cannot enter the prose unless the pipeline produced it, and the test suite asserts that every key either document quotes exists in the record, that the record did not come from a reduced-effort run, and — new here, because the previous record's shuffle count did not match the shipped default and nothing caught it — that **every statistical parameter in the record equals the shipped default**. +Beyond the paper, `flylab run` turns a declarative spec into a self-describing run directory whose manifest records the spec, code and library hashes, the seeds and two hashes per artifact; the same spec and seed reproduce it byte for byte (`docs/WORKFLOW.md`). `papers/results.json` carries every scalar quoted above with its value, rendered text, unit and producing step; two runs at one commit and seed produce identical values, runtimes excepted, and a test asserts it. Numeric placeholders in this template are drawn from that record. Tests check placeholder presence, the effort mode and recorded statistical defaults; those tests do not independently verify source-value attribution or reproduce the absent larger scale inputs. -The same wheel runs in a browser tab on Pyodide and is published as a static site, so these assays can be re-run by opening one link with nothing installed; `tests/test_browser_bridge.py` asserts **endpoint parity** — equal JSON, every readout, warning and provenance field, for a representative call to each route, which is route-level equality rather than a whole-analysis equality claim. MaleCNS v1.0 and FlyWire are CC-BY 4.0 from their consortia [5], [6] and must be cited before this software; licences are listed under Statements, and an archival Zenodo deposit will carry the DOI **[to be minted on acceptance]**. +The wheel also runs in a browser tab on Pyodide; `tests/test_browser_bridge.py` asserts **endpoint parity** — equal JSON, every readout, warning and provenance field, for a representative call to each route, which is route-level equality rather than a whole-analysis equality claim. The MaleCNS download page lists CC-BY; the FlyWire public release guidelines list CC BY-NC 4.0 [5], [6]. Both sources require attribution. The code release has no separate archival DOI. ## 4. Discussion -**What transfers.** *First*, the dependence verdict is a joint property of a prediction and the extract it is measured on, and the extract's recurrence predicts which way it goes. The same validated instrument, with no parameter changed, classifies {{dep_land_composition_dominated}} of {{dep_land_cells}} cells composition-dominated on a sparse extract (mean degree {{named_mean_degree}}) and {{dep_land_taste_composition_dominated}} of {{dep_land_taste_cells}} on a denser one (mean degree {{taste_motor_mean_degree}}), and across {{scale_study_n_cuts}} extracts the composition-dominated verdict survives on {{scale_study_n_composition_dominated}} — the in-star — while an extract with *fewer* nodes and twenty times the mean degree is already topology-dependent. The practical rule generalises to any pipeline of this shape: **a dependence result must be reported with the structural statistics of the extract it was measured on, and a negative measured on a sparse extract is not evidence about the network it was cut from.** Because the structural statistic that predicts the flip is computable before any permutation is run (T19, T27), this is a cheap check, not a research programme. +**Dependence interpretation.** The same procedure classifies {{dep_land_composition_dominated}} of {{dep_land_cells}} cells as composition-dominated on `named` and {{dep_land_taste_composition_dominated}} of {{dep_land_taste_cells}} on `taste_motor`, under their declared floors and test families. This is evidence that a dependence label is conditional on the cut, its construction and the runtime. The imported scale record adds descriptive observations across {{scale_study_n_cuts}} cuts, but missing large-cut inputs and changing permutation budgets prevent a separate reproduction of the complete scale ladder. The observed structural differences suggest hypotheses about recurrence; they do not identify recurrence as the cause of the label change. *Second*, typing the evidence changes results rather than documentation: refusing inadmissible rows removed a class of small, plausible, invented numbers no reader could have told apart from measurements, and grading a source's distance demoted a cross-species constant to a labelled proxy without changing its value. Stated as an invariant and checked exhaustively, this is a property of the pipeline rather than a convention its authors intend to follow. *Third*, two of the instruments were themselves broken in ways that flattered the conclusions, and both were found by pointing the instruments inward: a null that moved the quantity it claimed to preserve exactly, and a specification family whose alternatives never reached the engine its predicates ran on, so a uniform robustness rate was guaranteed by construction. Both failures share a signature worth naming — **a test whose result was determined by its own implementation rather than by the data** — and both were invisible to determinism checks, because a record can be perfectly reproducible and still be produced by a test that could not have returned another answer. The specification family and the variance decomposition then converge on one weak point: the direction of the headline effect is set by an asserted functional form, which is also the largest resolved factor in the variance budget and the top-ranked measurement. -**Encoded assumptions versus generated observations.** The separation is itself generated rather than asserted: an audit of the dependency chain behind one readout returns {{claims_chain_links}} links (T18), of which exactly {{claims_observed_phrase}} is a measurement of the system being simulated, {{claims_literature_derived}} reuses a measurement of something else, {{claims_model_assumption}} are asserted modelling choices and {{claims_computed}} are arithmetic on the links above. A companion decomposition — a different decomposition, not a repartition of those links — reports {{claims_facts_phrase}}, {{claims_inference_phrase}} and {{claims_unknown_phrase}}. That one link in nine is a measurement of the simulated system is worth stating plainly for any pipeline of this shape. +**Encoded assumptions versus generated observations.** An audit of the dependency chain behind one readout returns {{claims_chain_links}} links (T18): {{claims_observed_phrase}} measures the simulated system, {{claims_literature_derived}} reuses a measurement of something else, {{claims_model_assumption}} are asserted model choices and {{claims_computed}} are computational links. This is a trace for one readout, not a partition of biological evidence quality. -**Relation to adjacent work.** Our version of the [69], [70] result survives neither a change of extract (§3.2) nor a change of the runtime's normalisation (§3.3), and both models normalise or regionally average, so the parallel is at least as consistent with a shared modelling artefact as with a property of receptor-weighted network perturbation. That is itself a methodological observation: a result reproduced across two pipelines that share a normalisation convention has not been independently confirmed. +**Relation to adjacent work.** Similar model-output associations in receptor-map network models [68], [69] motivate comparison of normalisation and spatial aggregation. The current experiments do not establish a shared mechanism or a biological correspondence between those domains. **What the pipeline cannot support**, and these limits are themselves computing results about what such a pipeline can establish. It cannot predict a dose–response in an animal: with no pharmacokinetics available, the free concentration is an *input*, not a derived quantity. It cannot support a cross-species safety claim, nor a subunit-level claim where the sources used hybrid constructs. It cannot support a general statement about whether perturbation on networks of this kind needs topology — only a statement per extract with that extract's structure reported. And it cannot produce a live-animal result, which no code path may write. -**Five model-specific caveats** are given in full in §S9: an asserted sign table exempt from the specification family and unbounded by the label-permuting null; the load-bearing row normalisation; perturbations applied across transmitter-defined rather than receptor-defined synapses, which the expression layer bounds rather than corrects; hops-limited extracts, whose missing inhibition can flip the sign of any readout and whose cost §3.2 now quantifies; and agreement of direction only between the two runtimes. +**Model-specific caveats** are detailed in §S9: fixed asserted transmitter sign magnitudes, load-bearing row normalisation, perturbations applied across transmitter-defined rather than receptor-defined synapses, truncated extracts with missing inputs, and only qualitative direction agreement between the two runtimes. The gain-rule family does not vary the sign magnitudes or transmitter assignment model. -**Prospective, not pre-registered.** The hypotheses in §S5 are written before any experiment and labelled as software predictions with no live result; a mutable repository is not a registry, so they become pre-registered only once a tagged release is externally archived. Two of them predict a readout whose *sign* is a modelling choice over this paper's own specification family, and the supplement now says so beside them. +**Prospective hypotheses.** The hypotheses in §S5 are software predictions with no live result. They are not preregistered; a mutable repository is not a registry. The direction of two readouts changes within the tested specification family, as stated beside those hypotheses. ## 5. Conclusion -We set out to make a network simulation say which of its inputs its predictions depend on, and to evaluate that machinery rather than assert it. The method is four instruments — typed evidence propagation under a checked soundness invariant, input-dependence testing against an ordered ladder with an explicit equivalence margin and multiplicity control, an ablation ladder scored against a matched reference, and specification-family robustness with variance attribution — none specific to the domain we demonstrate them on. Pointed at its own results, the method establishes that the headline effect's direction is a property of an asserted functional form; that a layer-agreement score near 1 can be an algebraic consequence of the pipeline's own normalisation; and — the result we would most want another group to break — that **a dependence verdict obtained on one extract of a network can invert on another extract of the same network, with recurrence rather than size predicting which way**. A "this prediction does not need the structure" result is only as good as the extract it was measured on, and the statistic that says so is cheap to compute in advance. +This study implements typed admission of parameter rows, graph-degradation comparisons, ablation references and gain-rule sensitivity checks for two cuts of MaleCNS v1.0. It finds that some model-output directions depend on the asserted gain rule and that descriptive dependence labels differ between the two cuts. The source-value audit, missing inputs for larger saved scale rungs and absent circuit-level experimental validation limit biological interpretation. The graph-null and point-gap results are conditional on the declared extract, runtime, margins and component-test family. -What would change the picture is short and specific, and the variance budget rather than taste names it: a calibration of the parameter-to-gain transformation at a known synapse; a reconstruction-confidence re-analysis, which needs no new measurement; a full landscape on an extract that is neither an in-star nor hops-limited; and one live assay. Until then this is a description of a method, and of a tool that tests itself. +Next steps are direct calibration of the engagement-to-gain transformation, verification of the remaining source-to-parameter links, reconstruction-confidence analyses, reproducible larger-cut tests and an independent circuit-level assay. Until these are available, the results are a methodological case study and software-derived predictions. ## Statements -**AI assistance.** The author defined the research questions, methods and analyses; verified the source literature, checking each cited publication against the value attributed to it; executed and reviewed every analysis; and accepts full responsibility for the scientific content. AI assistance was used for software implementation and for drafting and editing prose. No data were generated, imputed, estimated or extrapolated by a language model; every number in the results, tables and figures was produced by the reproduction pipeline from committed artifacts and is keyed in `papers/results.json`. +**AI use.** OpenAI Codex assisted code implementation, computational reruns and checks, bibliography and source-value screening, figure preparation, and language editing. Claude Opus 5.5 and Grok supplied informal manuscript critiques. The author remains responsible for the methods, analyses, citations and final text. A row-level audit identified unresolved source-value links; these are reported as limitations. Numerical results are produced by the stated code and parameter table, not measured in living animals. **Animal research.** No animal experiments were performed and no live-animal data are reported. The `live_lab` field of every notebook produced by this pipeline is `null`. The protocols in `protocols/` are unexecuted drafts. -**Data and code availability.** Code: https://github.com/pinkysworld/FlyLab (MIT). Connectome data: official MaleCNS v1.0 and FlyWire releases, CC-BY 4.0, downloaded and not redistributed; derived cuts are committed under `data/derived/` with their citation strings. Literature datasets: `data/literature/`, each value carrying a DOI or PMID. +**Data and code availability.** Versioned analysis code and derived cuts are available at https://github.com/pinkysworld/FlyLab/tree/d0ce21d777000b38bb57d786c308a059c4430c95 (Git revision `d0ce21d777000b38bb57d786c308a059c4430c95`; MIT). The repository contains keyed results and the literature parameter library; source-to-value support remains incomplete for some numeric rows. The official MaleCNS v1.0 dataset (CC-BY) and FlyWire public release (CC BY-NC 4.0) were downloaded but are not redistributed. Literature datasets are under `data/literature/`. A separate archival DOI for the code has not been minted; the article DOI is assigned by the journal upon publication. + +**Funding.** This study received no funding. + +**Competing interests.** The author declares no competing interests. -**Competing interests.** None declared. **Funding.** None. +**Author contributions.** The sole author was responsible for conceptualization, methodology, software, formal analysis, interpretation and writing. + +**SDG alignment.** The audit method and reusable analysis code are relevant to SDG 9, particularly target 9.5 on scientific research and innovation. The study does not measure progress toward an SDG target. ## References -[1] J. Dempster, *RatCVS*, University of Strathclyde, Glasgow, UK. Simulated pithed-rat cardiovascular pharmacology teaching package. +[1] J. Dempster, *RatCVS: Rat Cardiovascular System Simulation*, GitHub repository. [Online]. Available: https://github.com/johndempster/RatCVS (accessed 23 Sep. 2026). -[2] S. Okabe *et al.*, "A web-based simulation of cardiovascular pharmacology for undergraduate teaching," *Eur. J. Pharmacol.*, vol. 997, art. 177618, 2025. +[2] S. Okabe, T. Goto, D. Hirayama and Y. Nishimura, "Virtual rat web: A versatile simulation tool for pharmacology education in a variety of settings," *Eur. J. Pharmacol.*, vol. 997, art. 177618, 2025. DOI 10.1016/j.ejphar.2025.177618. -[3] *PharmVR*: immersive virtual-reality pharmacology training, preprint, 2025. +[3] K. Thiyagarajan, "PharmVR—Immersive Virtual Reality based Animal Simulator Software," *Preprints.org*, concept paper, 2025. DOI 10.20944/preprints202507.0959.v1. Not peer reviewed. -[4] "Virtual simulation in pharmacology and physiology education: an evaluation," *J. Educ. Eval. Health Prof.*, 2020. +[4] C. C. Ezeala, "Integration of computer-simulated practical exercises into undergraduate medical pharmacology education at Mulungushi University, Zambia," *J. Educ. Eval. Health Prof.*, vol. 17, art. 8, 2020. DOI 10.3352/jeehp.2020.17.8. [5] S. Dorkenwald *et al.*, "Neuronal wiring diagram of an adult brain," *Nature*, vol. 634, pp. 124–138, 2024. DOI 10.1038/s41586-024-07558-y. -[6] HHMI Janelia FlyEM, University of Cambridge, MRC LMB and Google Research, *MaleCNS v1.0: a synapse-resolution connectome of the adult male* Drosophila *central nervous system*, 2026. CC-BY 4.0. - -[7] P. K. Shiu, G. R. Sterne, N. Spiller *et al.*, "A leaky integrate-and-fire computational model based on the connectome of the entire adult *Drosophila* brain reveals insights into sensorimotor processing," *Nature*, vol. 634, 2024. +[6] S. Berg *et al.*, "Sexual dimorphism in the complete *Drosophila* male central nervous system connectome," *Cell*, vol. 189, no. 18, pp. 5504–5526.e15, 2026. DOI 10.1016/j.cell.2026.08.015. Associated dataset: MaleCNS v1.0, https://male-cns.janelia.org/download/ (CC-BY 4.0). -[8] C. Bass *et al.*, "Mutation analysis and functional characterisation of insect nicotinic acetylcholine receptors," *Insect Biochem. Mol. Biol.*, vol. 36, pp. 86–96, 2006. +[7] P. K. Shiu, G. R. Sterne, N. Spiller *et al.*, "A *Drosophila* computational brain model reveals sensorimotor processing," *Nature*, vol. 634, pp. 210–219, 2024. DOI 10.1038/s41586-024-07763-9. -[9] H. Dederer *et al.*, "Selectivity of neonicotinoids at hybrid insect nicotinic acetylcholine receptors," *Insect Biochem. Mol. Biol.*, vol. 41, p. 51, 2011. +[8] H. Dederer, M. Werr and T. Ilg, "Differential sensitivity of *Ctenocephalides felis* and *Drosophila melanogaster* nicotinic acetylcholine receptor α1 and α2 subunits in recombinant hybrid receptors to nicotinoids and neonicotinoid insecticides," *Insect Biochem. Mol. Biol.*, vol. 41, no. 1, pp. 51–61, 2011. DOI 10.1016/j.ibmb.2010.09.012. -[10] FlyConnectome, *Neurotransmitter ground-truth repository for* Drosophila *connectomes*, 2024. +[9] FlyConnectome, *Drosophila Neurotransmitters: Neurotransmitter Ground Truth*, GitHub repository, 2024. [Online]. Available: https://github.com/flyconnectome/drosophila_neurotransmitters (accessed 23 Sep. 2026). -[11] P. Mineault, "Deconstructing viral fly simulations," technical note, 2026. +[10] P. Mineault, "Deconstructing viral fly sims," *The NeuroAI Archive*, 14 Sep. 2026. [Online]. Available: https://www.neuroai.science/p/are-flies-playing-beat-saber. Newsletter essay, not peer reviewed. -[12] Y. Deng *et al.*, "Chemoconnectomics: mapping chemical transmission in *Drosophila*," *Neuron*, vol. 101, pp. 876–893, 2019. +[11] B. Deng *et al.*, "Chemoconnectomics: Mapping Chemical Transmission in *Drosophila*," *Neuron*, vol. 101, no. 5, pp. 876–893.e4, 2019. DOI 10.1016/j.neuron.2019.01.045. -[13] C. Bass, A. M. Puinean, M. Andrews *et al.*, "Mutation of a nicotinic acetylcholine receptor β subunit is associated with resistance to neonicotinoid insecticides in the aphid *Myzus persicae*," *BMC Neurosci.*, vol. 12, art. 51, 2011. DOI 10.1186/1471-2202-12-51. PMID 21627790. +[12] C. Bass, A. M. Puinean, M. Andrews *et al.*, "Mutation of a nicotinic acetylcholine receptor β subunit is associated with resistance to neonicotinoid insecticides in the aphid *Myzus persicae*," *BMC Neurosci.*, vol. 12, art. 51, 2011. DOI 10.1186/1471-2202-12-51. PMID 21627790. -[14] R. H. ffrench-Constant, T. A. Rocheleau, J. C. Steichen and A. E. Chalmers, "A point mutation in a *Drosophila* GABA receptor confers insecticide resistance," *Nature*, vol. 363, pp. 449–451, 1993. DOI 10.1038/363449a0. +[13] R. H. ffrench-Constant, T. A. Rocheleau, J. C. Steichen and A. E. Chalmers, "A point mutation in a *Drosophila* GABA receptor confers insecticide resistance," *Nature*, vol. 363, pp. 449–451, 1993. DOI 10.1038/363449a0. -[15] W. T. Garrood, C. T. Zimmer, O. Gutbrod *et al.*, "Influence of the RDL A301S mutation in the brown planthopper *Nilaparvata lugens* on the activity of phenylpyrazole insecticides," *Pestic. Biochem. Physiol.*, vol. 142, pp. 1–8, 2017. DOI 10.1016/j.pestbp.2017.01.007. PMID 29107231. +[14] W. T. Garrood, C. T. Zimmer, O. Gutbrod *et al.*, "Influence of the RDL A301S mutation in the brown planthopper *Nilaparvata lugens* on the activity of phenylpyrazole insecticides," *Pestic. Biochem. Physiol.*, vol. 142, pp. 1–8, 2017. DOI 10.1016/j.pestbp.2017.01.007. PMID 29107231. -[16] A. M. Hosie, H. A. Baylis, S. D. Buckingham and D. B. Sattelle, "Actions of the insecticide fipronil on dieldrin-sensitive and -resistant GABA receptors of *Drosophila melanogaster*," *Br. J. Pharmacol.*, vol. 115, no. 6, pp. 909–912, 1995. DOI 10.1111/j.1476-5381.1995.tb15896.x. PMID 7582519. +[15] A. M. Hosie, H. A. Baylis, S. D. Buckingham and D. B. Sattelle, "Actions of the insecticide fipronil on dieldrin-sensitive and -resistant GABA receptors of *Drosophila melanogaster*," *Br. J. Pharmacol.*, vol. 115, no. 6, pp. 909–912, 1995. DOI 10.1111/j.1476-5381.1995.tb15896.x. PMID 7582519. -[17] A. M. Hosie and D. B. Sattelle, "Allosteric modulation of an expressed homo-oligomeric GABA-gated chloride channel of *Drosophila melanogaster*," *Br. J. Pharmacol.*, vol. 117, no. 6, pp. 1229–1237, 1996. DOI 10.1111/j.1476-5381.1996.tb16720.x. PMID 8882620. +[16] A. M. Hosie and D. B. Sattelle, "Allosteric modulation of an expressed homo-oligomeric GABA-gated chloride channel of *Drosophila melanogaster*," *Br. J. Pharmacol.*, vol. 117, no. 6, pp. 1229–1237, 1996. DOI 10.1111/j.1476-5381.1996.tb16720.x. PMID 8882620. -[18] D. H. Kwon, K. S. Yoon, J. M. Clark and S. H. Lee, "A point mutation in a glutamate-gated chloride channel confers abamectin resistance in the two-spotted spider mite, *Tetranychus urticae* Koch," *Insect Mol. Biol.*, vol. 19, pp. 583–591, 2010. DOI 10.1111/j.1365-2583.2010.01017.x. PMID 20522121. +[17] D. H. Kwon, K. S. Yoon, J. M. Clark and S. H. Lee, "A point mutation in a glutamate-gated chloride channel confers abamectin resistance in the two-spotted spider mite, *Tetranychus urticae* Koch," *Insect Mol. Biol.*, vol. 19, pp. 583–591, 2010. DOI 10.1111/j.1365-2583.2010.01017.x. PMID 20522121. -[19] G. Le Goff, A. Hamon, J.-B. Bergé and M. Amichot, "Resistance to fipronil in *Drosophila simulans*: influence of two point mutations in the RDL GABA receptor subunit," *J. Neurochem.*, vol. 92, pp. 1295–1305, 2005. DOI 10.1111/j.1471-4159.2004.02922.x. PMID 15748149. +[18] G. Le Goff, A. Hamon, J.-B. Bergé and M. Amichot, "Resistance to fipronil in *Drosophila simulans*: influence of two point mutations in the RDL GABA receptor subunit," *J. Neurochem.*, vol. 92, pp. 1295–1305, 2005. DOI 10.1111/j.1471-4159.2004.02922.x. PMID 15748149. -[20] P. Menozzi, M. A. Shi, A. Lougarre, Z. H. Tang and D. Fournier, "Mutations of acetylcholinesterase which confer insecticide resistance in *Drosophila melanogaster* populations," *BMC Evol. Biol.*, vol. 4, art. 4, 2004. DOI 10.1186/1471-2148-4-4. PMID 15018651. +[19] P. Menozzi, M. A. Shi, A. Lougarre, Z. H. Tang and D. Fournier, "Mutations of acetylcholinesterase which confer insecticide resistance in *Drosophila melanogaster* populations," *BMC Evol. Biol.*, vol. 4, art. 4, 2004. DOI 10.1186/1471-2148-4-4. PMID 15018651. -[21] A. Mutero, M. Pralavorio, J. M. Bride and D. Fournier, "Resistance-associated point mutations in insecticide-insensitive acetylcholinesterase," *Proc. Natl. Acad. Sci. USA*, vol. 91, no. 13, pp. 5922–5926, 1994. DOI 10.1073/pnas.91.13.5922. PMID 8016090. +[20] A. Mutero, M. Pralavorio, J. M. Bride and D. Fournier, "Resistance-associated point mutations in insecticide-insensitive acetylcholinesterase," *Proc. Natl. Acad. Sci. USA*, vol. 91, no. 13, pp. 5922–5926, 1994. DOI 10.1073/pnas.91.13.5922. PMID 8016090. -[22] P. N. R. Usherwood, H. Vais, B. P. S. Khambay, T. G. E. Davies and M. S. Williamson, "Sensitivity of the *Drosophila para* sodium channel to DDT is not lowered by the super-kdr mutation M918T on the IIS4–S5 linker that profoundly reduces sensitivity to permethrin and deltamethrin," *FEBS Lett.*, vol. 579, no. 28, pp. 6317–6325, 2005. DOI 10.1016/j.febslet.2005.09.096. PMID 16263118. +[21] P. N. R. Usherwood, H. Vais, B. P. S. Khambay, T. G. E. Davies and M. S. Williamson, "Sensitivity of the *Drosophila para* sodium channel to DDT is not lowered by the super-kdr mutation M918T on the IIS4–S5 linker that profoundly reduces sensitivity to permethrin and deltamethrin," *FEBS Lett.*, vol. 579, no. 28, pp. 6317–6325, 2005. DOI 10.1016/j.febslet.2005.09.096. PMID 16263118. -[23] H. Vais, M. S. Williamson, S. J. Goodson *et al.*, "Activation of *Drosophila* sodium channels promotes modification by deltamethrin: reductions in affinity caused by knock-down resistance mutations," *J. Gen. Physiol.*, vol. 115, no. 3, pp. 305–318, 2000. DOI 10.1085/jgp.115.3.305. PMID 10694259. +[22] H. Vais, M. S. Williamson, S. J. Goodson *et al.*, "Activation of *Drosophila* sodium channels promotes modification by deltamethrin: reductions in affinity caused by knock-down resistance mutations," *J. Gen. Physiol.*, vol. 115, no. 3, pp. 305–318, 2000. DOI 10.1085/jgp.115.3.305. PMID 10694259. -[24] X. Wang, A. M. Puinean, A. O. O'Reilly *et al.*, "Mutations on M3 helix of *Plutella xylostella* glutamate-gated chloride channel confer unequal resistance to abamectin by two different mechanisms," *Insect Biochem. Mol. Biol.*, vol. 86, pp. 50–57, 2017. DOI 10.1016/j.ibmb.2017.05.006. PMID 28576654. +[23] X. Wang, A. M. Puinean, A. O. O'Reilly *et al.*, "Mutations on M3 helix of *Plutella xylostella* glutamate-gated chloride channel confer unequal resistance to abamectin by two different mechanisms," *Insect Biochem. Mol. Biol.*, vol. 86, pp. 50–57, 2017. DOI 10.1016/j.ibmb.2017.05.006. PMID 28576654. -[25] H. G. Zhang, R. H. ffrench-Constant and M. B. Jackson, "A unique amino acid of the *Drosophila* GABA receptor with influence on drug sensitivity by two mechanisms," *J. Physiol.*, vol. 479, no. 1, pp. 65–75, 1994. DOI 10.1113/jphysiol.1994.sp020278. PMID 7527461. +[24] H. G. Zhang, R. H. ffrench-Constant and M. B. Jackson, "A unique amino acid of the *Drosophila* GABA receptor with influence on drug sensitivity by two mechanisms," *J. Physiol.*, vol. 479, no. 1, pp. 65–75, 1994. DOI 10.1113/jphysiol.1994.sp020278. PMID 7527461. -[26] A. Cartereau, C. Martin and S. H. Thany, "Neonicotinoid insecticides differently modulate acetylcholine-induced currents on mammalian α7 nicotinic acetylcholine receptors," *Br. J. Pharmacol.*, vol. 175, no. 11, pp. 1987–1998, 2018. DOI 10.1111/bph.14018. +[25] A. Cartereau, C. Martin and S. H. Thany, "Neonicotinoid insecticides differently modulate acetylcholine-induced currents on mammalian α7 nicotinic acetylcholine receptors," *Br. J. Pharmacol.*, vol. 175, no. 11, pp. 1987–1998, 2018. DOI 10.1111/bph.14018. -[27] L. M. Cole, R. A. Nicholson and J. E. Casida, "Action of phenylpyrazole insecticides at the GABA-gated chloride channel," *Pestic. Biochem. Physiol.*, vol. 46, no. 1, pp. 47–54, 1993. DOI 10.1006/pest.1993.1035. +[26] L. M. Cole, R. A. Nicholson and J. E. Casida, "Action of phenylpyrazole insecticides at the GABA-gated chloride channel," *Pestic. Biochem. Physiol.*, vol. 46, no. 1, pp. 47–54, 1993. DOI 10.1006/pest.1993.1035. -[28] G. R. Dawson, K. A. Wafford, A. Smith *et al.*, "Anticonvulsant and adverse effects of avermectin analogs in mice are mediated through the γ-aminobutyric acid A receptor," *J. Pharmacol. Exp. Ther.*, vol. 295, no. 3, pp. 1051–1060, 2000. PMID 11082440. +[27] G. R. Dawson, K. A. Wafford, A. Smith *et al.*, "Anticonvulsant and adverse effects of avermectin analogs in mice are mediated through the γ-aminobutyric acid A receptor," *J. Pharmacol. Exp. Ther.*, vol. 295, no. 3, pp. 1051–1060, 2000. PMID 11082440. DOI 10.1016/S0022-3565(24)39005-6. -[29] K. Lees, M. Musgaard, S. Suwanmanee *et al.*, "Actions of agonists, fipronil and ivermectin on the predominant in vivo splice and edit variant (RDLbd, I/V) of the *Drosophila* GABA receptor expressed in *Xenopus laevis* oocytes," *PLoS ONE*, vol. 9, no. 5, art. e97468, 2014. DOI 10.1371/journal.pone.0097468. +[28] K. Lees, M. Musgaard, S. Suwanmanee *et al.*, "Actions of agonists, fipronil and ivermectin on the predominant in vivo splice and edit variant (RDLbd, I/V) of the *Drosophila* GABA receptor expressed in *Xenopus laevis* oocytes," *PLoS ONE*, vol. 9, no. 5, art. e97468, 2014. DOI 10.1371/journal.pone.0097468. -[30] T. Perry, D. G. Heckel, J. A. McKenzie and P. Batterham, "Effects of mutations in *Drosophila* nicotinic acetylcholine receptor subunits on sensitivity to insecticides targeting nicotinic acetylcholine receptors," *Pestic. Biochem. Physiol.*, vol. 102, no. 1, pp. 56–60, 2012. DOI 10.1016/j.pestbp.2011.10.010. +[29] T. Perry, D. G. Heckel, J. A. McKenzie and P. Batterham, "Effects of mutations in *Drosophila* nicotinic acetylcholine receptor subunits on sensitivity to insecticides targeting nicotinic acetylcholine receptors," *Pestic. Biochem. Physiol.*, vol. 102, no. 1, pp. 56–60, 2012. DOI 10.1016/j.pestbp.2011.10.010. -[31] G. S. Ratra and J. E. Casida, "GABA receptor subunit composition relative to insecticide potency and selectivity," *Toxicol. Lett.*, vol. 122, no. 3, pp. 215–222, 2001. DOI 10.1016/s0378-4274(01)00366-6. PMID 11489356. +[30] G. S. Ratra and J. E. Casida, "GABA receptor subunit composition relative to insecticide potency and selectivity," *Toxicol. Lett.*, vol. 122, no. 3, pp. 215–222, 2001. DOI 10.1016/s0378-4274(01)00366-6. PMID 11489356. -[32] G. S. Ratra, S. G. Kamita and J. E. Casida, "Role of human GABA_A receptor β3 subunit in insecticide toxicity," *Toxicol. Appl. Pharmacol.*, vol. 172, no. 3, pp. 233–240, 2001. DOI 10.1006/taap.2001.9154. PMID 11312652. +[31] G. S. Ratra, S. G. Kamita and J. E. Casida, "Role of human GABA_A receptor β3 subunit in insecticide toxicity," *Toxicol. Appl. Pharmacol.*, vol. 172, no. 3, pp. 233–240, 2001. DOI 10.1006/taap.2001.9154. PMID 11312652. -[33] Q. Shan, J. L. Haddrill and J. W. Lynch, "Ivermectin, an unconventional agonist of the glycine receptor chloride channel," *J. Biol. Chem.*, vol. 276, no. 16, pp. 12556–12564, 2001. DOI 10.1074/jbc.M011264200. +[32] Q. Shan, J. L. Haddrill and J. W. Lynch, "Ivermectin, an unconventional agonist of the glycine receptor chloride channel," *J. Biol. Chem.*, vol. 276, no. 16, pp. 12556–12564, 2001. DOI 10.1074/jbc.M011264200. -[34] K. Tasman, S. A. Rands and J. J. L. Hodge, "The power of *Drosophila melanogaster* for modeling neonicotinoid effects on pollinators and identifying novel mechanisms," *Front. Physiol.*, vol. 12, art. 659440, 2021. DOI 10.3389/fphys.2021.659440. +[33] K. Tasman, S. A. Rands and J. J. L. Hodge, "The power of *Drosophila melanogaster* for modeling neonicotinoid effects on pollinators and identifying novel mechanisms," *Front. Physiol.*, vol. 12, art. 659440, 2021. DOI 10.3389/fphys.2021.659440. -[35] M. Tomizawa and J. E. Casida, "Selective toxicity of neonicotinoids attributable to specificity of insect and mammalian nicotinic receptors," *Annu. Rev. Entomol.*, vol. 48, pp. 339–364, 2003. DOI 10.1146/annurev.ento.48.091801.112731. PMID 12208819. +[34] M. Tomizawa and J. E. Casida, "Selective toxicity of neonicotinoids attributable to specificity of insect and mammalian nicotinic receptors," *Annu. Rev. Entomol.*, vol. 48, pp. 339–364, 2003. DOI 10.1146/annurev.ento.48.091801.112731. PMID 12208819. -[36] M. Tomizawa and J. E. Casida, "Neonicotinoid insecticide toxicology: mechanisms of selective action," *Annu. Rev. Pharmacol. Toxicol.*, vol. 45, pp. 247–268, 2005. DOI 10.1146/annurev.pharmtox.45.120403.095930. +[35] M. Tomizawa and J. E. Casida, "Neonicotinoid insecticide toxicology: mechanisms of selective action," *Annu. Rev. Pharmacol. Toxicol.*, vol. 45, pp. 247–268, 2005. DOI 10.1146/annurev.pharmtox.45.120403.095930. -[37] M. Tomizawa, D. L. Lee and J. E. Casida, "Neonicotinoid insecticides: molecular features conferring selectivity for insect versus mammalian nicotinic receptors," *J. Agric. Food Chem.*, vol. 48, no. 12, pp. 6016–6024, 2000. DOI 10.1021/jf000873c. PMID 11312774. +[36] M. Tomizawa, D. L. Lee and J. E. Casida, "Neonicotinoid insecticides: molecular features conferring selectivity for insect versus mammalian nicotinic receptors," *J. Agric. Food Chem.*, vol. 48, no. 12, pp. 6016–6024, 2000. DOI 10.1021/jf000873c. PMID 11312774. -[38] A. J. Wolstenholme, "Glutamate-gated chloride channels," *J. Biol. Chem.*, vol. 287, no. 48, pp. 40232–40238, 2012. DOI 10.1074/jbc.R112.406280. +[37] A. J. Wolstenholme, "Glutamate-gated chloride channels," *J. Biol. Chem.*, vol. 287, no. 48, pp. 40232–40238, 2012. DOI 10.1074/jbc.R112.406280. -[39] A. M. Allen, M. C. Neville, S. Birtles *et al.*, "A single-cell transcriptomic atlas of the adult *Drosophila* ventral nerve cord," *eLife*, vol. 9, art. e54074, 2020. DOI 10.7554/eLife.54074. PMID 32314735. +[38] A. M. Allen, M. C. Neville, S. Birtles *et al.*, "A single-cell transcriptomic atlas of the adult *Drosophila* ventral nerve cord," *eLife*, vol. 9, art. e54074, 2020. DOI 10.7554/eLife.54074. PMID 32314735. -[40] V. Croset, C. D. Treiber and S. Waddell, "Cellular diversity in the *Drosophila* midbrain revealed by single-cell transcriptomics," *eLife*, vol. 7, art. e34550, 2018. DOI 10.7554/eLife.34550. PMID 29671739. +[39] V. Croset, C. D. Treiber and S. Waddell, "Cellular diversity in the *Drosophila* midbrain revealed by single-cell transcriptomics," *eLife*, vol. 7, art. e34550, 2018. DOI 10.7554/eLife.34550. PMID 29671739. -[41] K. Davie, J. Janssens, D. Koldere *et al.*, "A single-cell transcriptome atlas of the aging *Drosophila* brain," *Cell*, vol. 174, no. 4, pp. 982–998.e20, 2018. DOI 10.1016/j.cell.2018.05.057. PMID 29909982. +[40] K. Davie, J. Janssens, D. Koldere *et al.*, "A single-cell transcriptome atlas of the aging *Drosophila* brain," *Cell*, vol. 174, no. 4, pp. 982–998.e20, 2018. DOI 10.1016/j.cell.2018.05.057. PMID 29909982. -[42] L. Enell, Y. Hamasaka, A. Kolodziejczyk and D. R. Nässel, "γ-Aminobutyric acid (GABA) signaling components in *Drosophila*: immunocytochemical localization of GABA_B receptors in relation to the GABA_A receptor subunit RDL and a vesicular GABA transporter," *J. Comp. Neurol.*, vol. 505, no. 1, pp. 18–31, 2007. DOI 10.1002/cne.21472. PMID 17729251. +[41] L. Enell, Y. Hamasaka, A. Kolodziejczyk and D. R. Nässel, "γ-Aminobutyric acid (GABA) signaling components in *Drosophila*: immunocytochemical localization of GABA_B receptors in relation to the GABA_A receptor subunit RDL and a vesicular GABA transporter," *J. Comp. Neurol.*, vol. 505, no. 1, pp. 18–31, 2007. DOI 10.1002/cne.21472. PMID 17729251. -[43] N. Konstantinides, K. Kapuralin, C. Fadil *et al.*, "Phenotypic convergence: distinct transcription factors regulate common terminal features," *Cell*, vol. 174, no. 3, pp. 622–635.e13, 2018. DOI 10.1016/j.cell.2018.05.021. PMID 29909983. +[42] N. Konstantinides, K. Kapuralin, C. Fadil *et al.*, "Phenotypic convergence: distinct transcription factors regulate common terminal features," *Cell*, vol. 174, no. 3, pp. 622–635.e13, 2018. DOI 10.1016/j.cell.2018.05.021. PMID 29909983. -[44] H. Li, J. Janssens, M. De Waegeneer *et al.*, "Fly Cell Atlas: a single-nucleus transcriptomic atlas of the adult fruit fly," *Science*, vol. 375, no. 6584, art. eabk2432, 2022. DOI 10.1126/science.abk2432. PMID 35239393. +[43] H. Li, J. Janssens, M. De Waegeneer *et al.*, "Fly Cell Atlas: a single-nucleus transcriptomic atlas of the adult fruit fly," *Science*, vol. 375, no. 6584, art. eabk2432, 2022. DOI 10.1126/science.abk2432. PMID 35239393. -[45] C. I. Bliss, "The toxicity of poisons applied jointly," *Ann. Appl. Biol.*, vol. 26, no. 3, pp. 585–615, 1939. DOI 10.1111/j.1744-7348.1939.tb06990.x. +[44] C. I. Bliss, "The toxicity of poisons applied jointly," *Ann. Appl. Biol.*, vol. 26, no. 3, pp. 585–615, 1939. DOI 10.1111/j.1744-7348.1939.tb06990.x. -[46] W. R. Greco, G. Bravo and J. C. Parsons, "The search for synergy: a critical review from a response surface perspective," *Pharmacol. Rev.*, vol. 47, no. 2, pp. 331–385, 1995. PMID 7568331. +[45] W. R. Greco, G. Bravo and J. C. Parsons, "The search for synergy: a critical review from a response surface perspective," *Pharmacol. Rev.*, vol. 47, no. 2, pp. 331–385, 1995. PMID 7568331. -[47] Q. Huang, Y. Deng, T. Zhan and Y. He, "Synergistic and antagonistic effects of piperonyl butoxide in fipronil-susceptible and resistant rice stem borers, *Chilo suppressalis*," *J. Insect Sci.*, vol. 10, art. 182, 2010. DOI 10.1673/031.010.14142. PMID 21062143. +[46] Q. Huang, Y. Deng, T. Zhan and Y. He, "Synergistic and antagonistic effects of piperonyl butoxide in fipronil-susceptible and resistant rice stem borrers [sic], *Chilo suppressalis*," *J. Insect Sci.*, vol. 10, art. 182, 2010. DOI 10.1673/031.010.14142. PMID 21062143. -[48] S. Loewe and H. Muischnek, "Über Kombinationswirkungen: 1. Mitteilung: Hilfsmittel der Fragestellung," *Naunyn-Schmiedebergs Arch. Exp. Pathol. Pharmakol.*, vol. 114, pp. 313–326, 1926. DOI 10.1007/BF01952257. +[47] S. Loewe and H. Muischnek, "Über Kombinationswirkungen: 1. Mitteilung: Hilfsmittel der Fragestellung," *Naunyn-Schmiedebergs Arch. Exp. Pathol. Pharmakol.*, vol. 114, pp. 313–326, 1926. DOI 10.1007/BF01952257. -[49] S. Loewe, "The problem of synergism and antagonism of combined drugs," *Arzneimittelforschung*, vol. 3, no. 6, pp. 285–290, 1953. PMID 13081480. +[48] S. Loewe, "The problem of synergism and antagonism of combined drugs," *Arzneimittelforschung*, vol. 3, no. 6, pp. 285–290, 1953. PMID 13081480. -[50] D. Philippou, V. Borzatta, E. Capparella *et al.*, "The use of substituted alkynyl phenoxy derivatives of piperonyl butoxide to control insecticide-resistant pests," *Pest Manag. Sci.*, vol. 72, no. 10, pp. 1946–1950, 2016. DOI 10.1002/ps.4234. PMID 26800141. +[49] D. Philippou, V. Borzatta, E. Capparella *et al.*, "The use of substituted alkynyl phenoxy derivatives of piperonyl butoxide to control insecticide-resistant pests," *Pest Manag. Sci.*, vol. 72, no. 10, pp. 1946–1950, 2016. DOI 10.1002/ps.4234. PMID 26800141. -[51] Y. C. Zhu, J. Yao, J. Adamczyk and R. Luttrell, "Synergistic toxicity and physiological impact of imidacloprid alone and binary mixtures with seven representative pesticides on honey bee (*Apis mellifera*)," *PLoS ONE*, vol. 12, no. 5, art. e0176837, 2017. DOI 10.1371/journal.pone.0176837. PMID 28467462. +[50] Y. C. Zhu, J. Yao, J. Adamczyk and R. Luttrell, "Synergistic toxicity and physiological impact of imidacloprid alone and binary mixtures with seven representative pesticides on honey bee (*Apis mellifera*)," *PLoS ONE*, vol. 12, no. 5, art. e0176837, 2017. DOI 10.1371/journal.pone.0176837. PMID 28467462. -[52] P. J. Daborn, J. L. Yen, M. R. Bogwitz *et al.*, "A single P450 allele associated with insecticide resistance in *Drosophila*," *Science*, vol. 297, no. 5590, pp. 2253–2256, 2002. DOI 10.1126/science.1074170. PMID 12351787. +[51] P. J. Daborn, J. L. Yen, M. R. Bogwitz *et al.*, "A single P450 allele associated with insecticide resistance in *Drosophila*," *Science*, vol. 297, no. 5590, pp. 2253–2256, 2002. DOI 10.1126/science.1074170. PMID 12351787. -[53] D. G. Folk, C. Han and T. J. Bradley, "Water acquisition and partitioning in *Drosophila melanogaster*: effects of selection for desiccation-resistance," *J. Exp. Biol.*, vol. 204, no. 19, pp. 3323–3331, 2001. DOI 10.1242/jeb.204.19.3323. PMID 11606606. +[52] D. G. Folk, C. Han and T. J. Bradley, "Water acquisition and partitioning in *Drosophila melanogaster*: effects of selection for desiccation-resistance," *J. Exp. Biol.*, vol. 204, no. 19, pp. 3323–3331, 2001. DOI 10.1242/jeb.204.19.3323. PMID 11606606. -[54] R. Fusetto, S. Denecke, T. Perry, R. A. J. O'Hair and P. Batterham, "Partitioning the roles of CYP6G1 and gut microbes in the metabolism of the insecticide imidacloprid in *Drosophila melanogaster*," *Sci. Rep.*, vol. 7, art. 11339, 2017. DOI 10.1038/s41598-017-09800-2. PMID 28900131. +[53] R. Fusetto, S. Denecke, T. Perry, R. A. J. O'Hair and P. Batterham, "Partitioning the roles of CYP6G1 and gut microbes in the metabolism of the insecticide imidacloprid in *Drosophila melanogaster*," *Sci. Rep.*, vol. 7, art. 11339, 2017. DOI 10.1038/s41598-017-09800-2. PMID 28900131. -[55] N. Joußen, D. G. Heckel, M. Haas, I. Schuphan and B. Schmidt, "Metabolism of imidacloprid and DDT by P450 CYP6G1 expressed in cell cultures of *Nicotiana tabacum*," *Pest Manag. Sci.*, vol. 64, no. 1, pp. 65–73, 2008. DOI 10.1002/ps.1472. PMID 17912692. +[54] N. Joußen, D. G. Heckel, M. Haas, I. Schuphan and B. Schmidt, "Metabolism of imidacloprid and DDT by P450 CYP6G1 expressed in cell cultures of *Nicotiana tabacum* suggests detoxification of these insecticides in *Cyp6g1*-overexpressing strains of *Drosophila melanogaster*, leading to resistance," *Pest Manag. Sci.*, vol. 64, no. 1, pp. 65–73, 2008. DOI 10.1002/ps.1472. PMID 17912692. -[56] S. C. Piyankarage, H. Augustin, Y. Grosjean, D. E. Featherstone and S. A. Shippy, "Hemolymph amino acid analysis of individual *Drosophila* larvae," *Anal. Chem.*, vol. 80, no. 4, pp. 1201–1207, 2008. DOI 10.1021/ac701785z. +[55] S. C. Piyankarage, H. Augustin, Y. Grosjean, D. E. Featherstone and S. A. Shippy, "Hemolymph amino acid analysis of individual *Drosophila* larvae," *Anal. Chem.*, vol. 80, no. 4, pp. 1201–1207, 2008. DOI 10.1021/ac701785z. -[57] S. C. Piyankarage, D. E. Featherstone and S. A. Shippy, "Nanoliter hemolymph sampling and analysis of individual adult *Drosophila melanogaster*," *Anal. Chem.*, vol. 84, no. 10, pp. 4460–4466, 2012. DOI 10.1021/ac3002319. +[56] S. C. Piyankarage, D. E. Featherstone and S. A. Shippy, "Nanoliter hemolymph sampling and analysis of individual adult *Drosophila melanogaster*," *Anal. Chem.*, vol. 84, no. 10, pp. 4460–4466, 2012. DOI 10.1021/ac3002319. -[58] S. Suchail, L. Debrauwer and L. P. Belzunces, "Metabolism of imidacloprid in *Apis mellifera*," *Pest Manag. Sci.*, vol. 60, no. 3, pp. 291–296, 2004. DOI 10.1002/ps.772. PMID 15025241. +[57] S. Suchail, L. Debrauwer and L. P. Belzunces, "Metabolism of imidacloprid in *Apis mellifera*," *Pest Manag. Sci.*, vol. 60, no. 3, pp. 291–296, 2004. DOI 10.1002/ps.772. PMID 15025241. -[59] S. Suchail, G. De Sousa, R. Rahmani and L. P. Belzunces, "In vivo distribution and metabolisation of ¹⁴C-imidacloprid in different compartments of *Apis mellifera* L.," *Pest Manag. Sci.*, vol. 60, no. 11, pp. 1056–1062, 2004. DOI 10.1002/ps.895. +[58] S. Suchail, G. De Sousa, R. Rahmani and L. P. Belzunces, "In vivo distribution and metabolisation of ¹⁴C-imidacloprid in different compartments of *Apis mellifera* L.," *Pest Manag. Sci.*, vol. 60, no. 11, pp. 1056–1062, 2004. DOI 10.1002/ps.895. -[60] W. Cao, L. Song, J. Cheng *et al.*, "An automated rapid iterative negative geotaxis assay for analyzing adult climbing behavior in a *Drosophila* model of neurodegeneration," *J. Vis. Exp.*, no. 127, art. 56507, 2017. DOI 10.3791/56507. PMID 28931001. +[59] W. Cao, L. Song, J. Cheng *et al.*, "An automated rapid iterative negative geotaxis assay for analyzing adult climbing behavior in a *Drosophila* model of neurodegeneration," *J. Vis. Exp.*, no. 127, art. 56507, 2017. DOI 10.3791/56507. PMID 28931001. -[61] M. Ö. Çevik and A. Erden, "The course of habituation of the proboscis extension reflex can be predicted by sucrose responsiveness in *Drosophila*," *PLoS ONE*, vol. 7, no. 6, art. e39863, 2012. DOI 10.1371/journal.pone.0039863. PMID 22761915. +[60] M. Ö. Çevik and A. Erden, "The course of habituation of the proboscis extension reflex can be predicted by sucrose responsiveness in *Drosophila*," *PLoS ONE*, vol. 7, no. 6, art. e39863, 2012. DOI 10.1371/journal.pone.0039863. PMID 22761915. -[62] W. Chen, X. Gu, Y. T. Yang, P. Batterham and T. Perry, "Dual nicotinic acetylcholine receptor subunit gene knockouts reveal limits to functional redundancy," *Pestic. Biochem. Physiol.*, vol. 184, art. 105118, 2022. DOI 10.1016/j.pestbp.2022.105118. PMID 35715057. +[61] W. Chen, X. Gu, Y. T. Yang, P. Batterham and T. Perry, "Dual nicotinic acetylcholine receptor subunit gene knockouts reveal limits to functional redundancy," *Pestic. Biochem. Physiol.*, vol. 184, art. 105118, 2022. DOI 10.1016/j.pestbp.2022.105118. PMID 35715057. -[63] A. S. French, M. J. Sellier, M. Ali Agha *et al.*, "Dual mechanism for bitter avoidance in *Drosophila*," *J. Neurosci.*, vol. 35, no. 9, pp. 3990–4004, 2015. DOI 10.1523/JNEUROSCI.1312-14.2015. PMID 25740527. +[62] A. S. French, M. J. Sellier, M. Ali Agha *et al.*, "Dual mechanism for bitter avoidance in *Drosophila*," *J. Neurosci.*, vol. 35, no. 9, pp. 3990–4004, 2015. DOI 10.1523/JNEUROSCI.1312-14.2015. PMID 25740527. -[64] J. W. Gargano, I. Martin, P. Bhandari and M. S. Grotewiel, "Rapid iterative negative geotaxis (RING): a new method for assessing age-related locomotor decline in *Drosophila*," *Exp. Gerontol.*, vol. 40, no. 5, pp. 386–395, 2005. DOI 10.1016/j.exger.2005.02.005. PMID 15919590. +[63] J. W. Gargano, I. Martin, P. Bhandari and M. S. Grotewiel, "Rapid iterative negative geotaxis (RING): a new method for assessing age-related locomotor decline in *Drosophila*," *Exp. Gerontol.*, vol. 40, no. 5, pp. 386–395, 2005. DOI 10.1016/j.exger.2005.02.005. PMID 15919590. -[65] H. Y. Lee, L. Zhou, S. Ghanta *et al.*, "Mechanisms of naturally evolved ethanol resistance in *Drosophila melanogaster*," *J. Exp. Biol.*, 2014. DOI 10.1242/jeb.110510. +[64] J. D. Fry, "Mechanisms of naturally evolved ethanol resistance in *Drosophila melanogaster*," *J. Exp. Biol.*, vol. 217, no. 22, pp. 3996–4003, 2014. DOI 10.1242/jeb.110510. -[66] F. Martelli, Z. Zhongyuan, J. Wang *et al.*, "Low doses of the neonicotinoid insecticide imidacloprid induce ROS triggering neurological and metabolic impairments in *Drosophila*," *Proc. Natl. Acad. Sci. USA*, vol. 117, no. 41, pp. 25840–25850, 2020. DOI 10.1073/pnas.2011828117. PMID 32989137. +[65] F. Martelli, Z. Zhongyuan, J. Wang *et al.*, "Low doses of the neonicotinoid insecticide imidacloprid induce ROS triggering neurological and metabolic impairments in *Drosophila*," *Proc. Natl. Acad. Sci. USA*, vol. 117, no. 41, pp. 25840–25850, 2020. DOI 10.1073/pnas.2011828117. PMID 32989137. -[67] M. J. Burton, I. R. Mellor, I. R. Duce *et al.*, "Differential resistance of insect sodium channels with kdr mutations to deltamethrin, permethrin and DDT," *Insect Biochem. Mol. Biol.*, vol. 41, no. 9, pp. 723–732, 2011. DOI 10.1016/j.ibmb.2011.05.004. +[66] M. J. Burton, I. R. Mellor, I. R. Duce *et al.*, "Differential resistance of insect sodium channels with kdr mutations to deltamethrin, permethrin and DDT," *Insect Biochem. Mol. Biol.*, vol. 41, no. 9, pp. 723–732, 2011. DOI 10.1016/j.ibmb.2011.05.004. -[68] A. A. Lazar, T. Liu, M. K. Turkcan and Y. Zhou, "Accelerating with FlyBrainLab the discovery of the functional logic of the *Drosophila* brain in the connectomic and synaptomic era," *eLife*, vol. 10, art. e62362, 2021. DOI 10.7554/eLife.62362. PMID 33616035. +[67] A. A. Lazar, T. Liu, M. K. Turkcan and Y. Zhou, "Accelerating with FlyBrainLab the discovery of the functional logic of the *Drosophila* brain in the connectomic and synaptomic era," *eLife*, vol. 10, art. e62362, 2021. DOI 10.7554/eLife.62362. PMID 33616035. -[69] I. Mindlin, R. Herzog, L. Belloli *et al.*, "Whole brain modelling for simulating pharmacological interventions on patients with disorders of consciousness," *Commun. Biol.*, vol. 7, art. 1176, 2024. DOI 10.1038/s42003-024-06852-9. PMID 39300281. +[68] I. Mindlin, R. Herzog, L. Belloli *et al.*, "Whole brain modelling for simulating pharmacological interventions on patients with disorders of consciousness," *Commun. Biol.*, vol. 7, art. 1176, 2024. DOI 10.1038/s42003-024-06852-9. PMID 39300281. -[70] G. Deco, J. Cruzat, J. Cabral *et al.*, "Whole-brain multimodal neuroimaging model using serotonin receptor maps explains non-linear functional effects of LSD," *Curr. Biol.*, vol. 28, no. 19, pp. 3065–3074.e6, 2018. DOI 10.1016/j.cub.2018.07.083. PMID 30270185. +[69] G. Deco, J. Cruzat, J. Cabral *et al.*, "Whole-brain multimodal neuroimaging model using serotonin receptor maps explains non-linear functional effects of LSD," *Curr. Biol.*, vol. 28, no. 19, pp. 3065–3074.e6, 2018. DOI 10.1016/j.cub.2018.07.083. PMID 30270185. -[71] A. Saltelli, P. Annoni, I. Azzini, F. Campolongo, M. Ratto and S. Tarantola, "Variance based sensitivity analysis of model output: design and estimator for the total sensitivity index," *Comput. Phys. Commun.*, vol. 181, no. 2, pp. 259–270, 2010. DOI 10.1016/j.cpc.2009.09.018. +[70] A. Saltelli, P. Annoni, I. Azzini, F. Campolongo, M. Ratto and S. Tarantola, "Variance based sensitivity analysis of model output: design and estimator for the total sensitivity index," *Comput. Phys. Commun.*, vol. 181, no. 2, pp. 259–270, 2010. DOI 10.1016/j.cpc.2009.09.018. -[72] M. J. W. Jansen, "Analysis of variance designs for model output," *Comput. Phys. Commun.*, vol. 117, no. 1–2, pp. 35–43, 1999. DOI 10.1016/S0010-4655(98)00154-4. +[71] M. J. W. Jansen, "Analysis of variance designs for model output," *Comput. Phys. Commun.*, vol. 117, no. 1–2, pp. 35–43, 1999. DOI 10.1016/S0010-4655(98)00154-4. -[73] T. Ishigami and T. Homma, "An importance quantification technique in uncertainty analysis for computer models," in *Proc. 1st Int. Symp. Uncertainty Modeling and Analysis (ISUMA)*, 1990, pp. 398–403. DOI 10.1109/ISUMA.1990.151285. +[72] T. Ishigami and T. Homma, "An importance quantification technique in uncertainty analysis for computer models," in *Proc. 1st Int. Symp. Uncertainty Modeling and Analysis (ISUMA)*, 1990, pp. 398–403. DOI 10.1109/ISUMA.1990.151285. -[74] S. Steegen, F. Tuerlinckx, A. Gelman and W. Vanpaemel, "Increasing transparency through a multiverse analysis," *Perspect. Psychol. Sci.*, vol. 11, no. 5, pp. 702–712, 2016. DOI 10.1177/1745691616658637. PMID 27694465. +[73] S. Steegen, F. Tuerlinckx, A. Gelman and W. Vanpaemel, "Increasing transparency through a multiverse analysis," *Perspect. Psychol. Sci.*, vol. 11, no. 5, pp. 702–712, 2016. DOI 10.1177/1745691616658637. PMID 27694465. -[75] U. Simonsohn, J. P. Simmons and L. D. Nelson, "Specification curve analysis," *Nat. Hum. Behav.*, vol. 4, pp. 1208–1214, 2020. DOI 10.1038/s41562-020-0912-z. PMID 32719546. +[74] U. Simonsohn, J. P. Simmons and L. D. Nelson, "Specification curve analysis," *Nat. Hum. Behav.*, vol. 4, pp. 1208–1214, 2020. DOI 10.1038/s41562-020-0912-z. PMID 32719546. -[76] R. A. Howard, "Information value theory," *IEEE Trans. Syst. Sci. Cybern.*, vol. 2, no. 1, pp. 22–26, 1966. DOI 10.1109/TSSC.1966.300074. +[75] R. A. Howard, "Information value theory," *IEEE Trans. Syst. Sci. Cybern.*, vol. 2, no. 1, pp. 22–26, 1966. DOI 10.1109/TSSC.1966.300074. -[77] M. Tomizawa, B. Latli and J. E. Casida, "Novel neonicotinoid-agarose affinity column for *Drosophila* and *Musca* nicotinic acetylcholine receptors," *J. Neurochem.*, vol. 67, no. 4, pp. 1669–1676, 1996. DOI 10.1046/j.1471-4159.1996.67041669.x. PMID 8858952. +[76] M. Tomizawa, B. Latli and J. E. Casida, "Novel neonicotinoid-agarose affinity column for *Drosophila* and *Musca* nicotinic acetylcholine receptors," *J. Neurochem.*, vol. 67, no. 4, pp. 1669–1676, 1996. DOI 10.1046/j.1471-4159.1996.67041669.x. PMID 8858952. --- diff --git a/papers/STATUS.md b/papers/STATUS.md index 4039b52..ac66dfc 100644 --- a/papers/STATUS.md +++ b/papers/STATUS.md @@ -43,7 +43,7 @@ python -m pytest -q # must be green These are in the manuscript on purpose and must not be removed to make it look better: -- **The central RQ2 claim reversed, and the scaling study says what it tracks.** On the in-star `named` cut the landscape gives 11 topology-dependent / 42 composition-dominated / 31 no-effect of 84; repeated unchanged on `taste_motor` it gives 54 / 0 / 30. Across a ladder of 7 extracts (`scale_1k` … `scale_50k` plus the two committed cuts) the composition-dominated verdict survives on **exactly one** — the in-star — and `scale_1k` has *fewer* nodes (1000 vs 1126) with mean degree 22.9 vs 1.21 and is already topology-dependent. **Recurrence, not size, predicts the inversion.** At the denser middle rungs (5k, 10k) imidacloprid is distinguishable from *every* null. The top rungs run at n = 20 (resolution 0.048), so they confirm rather than establish the settled verdict. +- **The central RQ2 claim reversed; substrate structure is associated with the verdict.** On the in-star `named` cut the landscape gives 11 topology-dependent / 42 composition-dominated / 31 no-effect of 84; repeated unchanged on `taste_motor` it gives 54 / 0 / 30. Across a ladder of 7 extracts (`scale_1k` … `scale_50k` plus the two committed cuts) the composition-dominated verdict occurs on **exactly one** — the in-star — and `scale_1k` has *fewer* nodes (1000 vs 1126) with mean degree 22.9 vs 1.21 and is already topology-dependent. These observations do not isolate recurrence or any other structural feature as causal. At the denser middle rungs (5k, 10k) imidacloprid is distinguishable from the three ranked edge/wiring nulls; the plain transmitter-label permutation is a separate joint target-and-sign check. The top rungs run at n = 20 (resolution 0.048), so they confirm rather than establish the saved verdict. - Imidacloprid's `named` mean-rate effect is **composition-dominated** at n = 1000 — equivalent within tolerance for rewire_degree_preserving, sign_permute_weight_matched, weight_permute and merely indeterminate for sign_permute against a margin of 0.332 Hz. Fipronil is topology-dependent there (p = 0.006) but does **not** survive correction inside the exploratory landscape (q = 0.050). - The fipronil bitter-veto arm is a **properly powered negative** at n = 300 (all p > 0.05, |z| < 0.25). - **The suppression conclusion is specification-dependent**: retained by 15/25 specifications, reversed by all 10 monotone ones. And the previous 25/25 topology result was **vacuous** — 6 shuffles could not reject, and the specification never reached the null engine. diff --git a/papers/SUBMISSION_CHECKLIST.md b/papers/SUBMISSION_CHECKLIST.md index 45bf195..30ce651 100644 --- a/papers/SUBMISSION_CHECKLIST.md +++ b/papers/SUBMISSION_CHECKLIST.md @@ -14,7 +14,7 @@ python -m pytest -q # must be green IJRC is a computing research journal and the manuscript is written as one. The lead is the general problem — *a simulation built on a measured network and literature parameters cannot say which of its inputs its predictions depend on* — and the contribution is four transferable instruments: typed evidence propagation under a checked soundness invariant; input-dependence testing against an ordered ladder with an equivalence margin and multiplicity control; an ablation ladder scored against a matched reference; and specification-family robustness with variance attribution. §3 evaluates **the instruments**. The fly pharmacology is the demonstration domain, with its census, readouts and literature concordance as supporting material in §S6 and §S10. -The headline is methodological: **a dependence verdict obtained on one extract of a network can invert on another extract of the same network, and the extract's recurrence — not its size — predicts which way** (§3.2, T27). +The result is methodological: two structurally different extracts yield different descriptive dependence-label counts. Density, composition, extraction rule and recurrent paths vary together; the comparison does not identify recurrence or network size as the cause. The larger-scale study is an imported record whose 5k–50k input cuts are absent from this review package. Internal review history lives in `papers/reviews/`. It is archived critique, not journal correspondence: the corrections it forced are in the manuscript and the code, and a formal point-by-point response belongs to an actual resubmission against real referees' numbered comments. @@ -22,38 +22,40 @@ Internal review history lives in `papers/reviews/`. It is archived critique, not | Item | State | Where | |---|---|---| -| Title, author block, affiliations | **missing** — authors are "FlyLab contributors" | needs real names, ORCIDs, affiliations, a corresponding author and an email | +| Title, author block, affiliations | prepared on a separate editorial title page; withheld from the blind manuscript | Michél Nguyen, University of the People, ORCID and correspondence details | | Abstract | done — computing problem first, the methodological headline, the central limitation | draft §Abstract | | Index terms / keywords | done | draft §Index terms | -| Numbered IEEE-style references with DOIs | done — 77 entries | draft §References | +| Numbered references | 76 source-paper entries; DOI and PMID identities checked, assay-level support remains incomplete | draft §References and review audit | | In-text citation of every reference | check with the grep in §D | — | | Figures, 300 dpi, captions | done — 16 figures (F16 is the instrument validation) | `papers/figures/`, `figures/captions.md` | | Tables | done — T0–T27 as CSV and Markdown | `papers/tables/` | | Word count | done — inside the 5000–7000 target, generated and test-enforced | `values.paper_words_body` | | Required section structure | done | draft §1–§5 | -| Supplementary material | done — S1–S10 | `papers/SUPPLEMENT.md` | -| Data / code / ethics / AI statements | done | draft §Statements | -| Archival DOI | **placeholder** — Zenodo deposit not created | draft §3.7, `CITATION.cff` | -| Submission format (LaTeX/Word template) | **not done** — the draft is Markdown | convert once the venue's template is confirmed | +| Supplementary material | S1–S11 prepared, with cited figures and tables in a companion archive | `papers/SUPPLEMENT.md` and separate review files | +| Funding / competing interests | author declares no funding and no competing interests | draft §Statements and editorial title page | +| Author contributions / AI / ethics / SDG mapping | supplied; SDG 9.5 is a relevance mapping, not an impact result | draft §Statements and editorial title page | +| Data and code | numerical code at a commit-specific GitHub URL; masked review archive supplied | draft §Statements and review package | +| DOI | IJRC assigns the article DOI after publication; a separate code archive DOI is optional and currently absent | draft §Statements | +| Submission format | blinded Word manuscript, separate title page and formatted supplement prepared | local review package | | Cover letter | **not written** | — | | Suggested reviewers | **not chosen** | — | -## B. What is still missing for a **results** paper +## B. Work not claimed as completed -1. **One live *Drosophila* assay.** Climbing first: a complete published protocol and control statistics exist. Until a real table is imported by hand into `live_lab`, H1–H7 stay prospective software predictions. -2. **The one fitted parameter.** The variance budget names it: the engagement→gain transformation, not a potency value, which the model cannot see at a saturating dose. -3. **Better power at the top of the scale ladder.** The ladder settles the verdict, but its 25k/50k rungs run at n = 20 (resolution 0.048), where a rejection is the smallest the test can express. More permutations there, or a cheaper null, would turn confirmation into evidence. -4. **Expression coverage above its current value**, with adult motor-neuron receptor expression sourced or the bound declared permanent. -5. **An externally archived release**, without which "prospective" may not become "pre-registered". +1. **Biological validation.** No live *Drosophila* assay was performed. H1–H7 remain prospective software predictions, not empirical effects. +2. **Engagement-to-gain calibration.** The transformation and five transmitter-sign magnitudes remain assumed model inputs; the reported specification family does not vary all of them. +3. **Independent large-scale reproduction.** The 5k–50k cuts and their inputs are absent, so ST27 is an imported exploratory record. Its low permutation counts also limit resolution. +4. **Assay-level library provenance.** Bibliographic identity does not establish the potency value used in every numeric receptor row. +5. **Preregistration.** Neither a public repository nor a later code DOI can retrospectively preregister H1–H7; a dated, frozen experimental protocol would be needed before a future experiment. ## C. Known weaknesses a reviewer will find first -1. **The central RQ2 result reverses between the two cuts.** Answer: that *is* the result, it is measured rather than argued, the instrument is validated against planted ground truth (F16, T22) so the reversal is not an artefact of power, the structural difference is reported in advance (T19), and the scaling study over seven extracts (T27) identifies recurrence rather than size as what predicts it. -2. **The top rungs of the scale ladder run at n = 20.** Answer: stated in §3.2 and §S2.4 with the resolution (0.048 against α = 0.05) — those rungs confirm a verdict already settled at better-powered rungs, and the paper says they do not establish it. +1. **The two cuts yield different labels.** The manuscript reports descriptive differences and a limited planted-cycle positive control; it does not attribute the contrast causally to recurrence or claim general power for the paper endpoint. +2. **The top rungs of the saved scale study run at n = 20.** Their minimum attainable probability is 0.048, and the input cuts are unavailable here. The study is exploratory, not independent confirmation. 3. **The composition-versus-full correlation is nearly an algebraic identity.** Answer: stated as such, with a matched reference distribution (T24) and a normalisation sweep (T25); the generalisation drawn from it is withdrawn. -4. **The headline circuit result's direction depends on the gain rule.** Answer: a first-class result (§3.4), quantified over the prespecified specification family, and nominated by the variance budget as the measurement to make next. +4. **One model-output direction depends on the gain rule.** The tested family is reported in §3.4 and §S1.1; fixed sign magnitudes and transmitter assignment remain outside it. 5. **The gain rules and five transmitter signs are asserted coefficients.** Answer: published (T2, T21) and named as such; the paper says which of its own claims that costs it. -6. **No independent out-of-sample validation.** Answer: stated in §2.5 and §S10, with shared-source rank comparisons flagged by the pipeline. The ground-truth experiment validates the instrument, not the domain model, and the paper says so. +6. **No independent out-of-sample validation.** This is stated in §2.5 and §S10, with detectable shared-source rank comparisons flagged by the pipeline. The planted-cycle test is a limited synthetic positive control; it does not validate the domain model or estimate general detection power. 7. **Two of the instruments were broken.** Answer: found by us, published in §3.4, §S1.1 and §S2, and both fixed with the corrected numbers in the text. Neither failure was detectable by a determinism check, which is itself reported as a finding. 8. **`--fast` is not the paper.** Answer: one test refuses a committed record from a fast run, another asserts the record's statistical knobs equal the shipped defaults. @@ -99,9 +101,10 @@ python -m pytest -q tests/test_reproduce.py ## E. Final pass before sending -- [ ] Real authors, affiliations, ORCIDs, corresponding author. -- [ ] Zenodo deposit created; DOI substituted for the placeholder in the draft and in `CITATION.cff`; only then may the predictions be called pre-registered. -- [ ] Tag the release; confirm `git_sha` in `papers/results.json` matches the tag. +- [x] Author, affiliation, ORCID and correspondence details are on the separate editorial title page. +- [x] Funding, competing interests, AI use, contributions, ethics, data/code and SDG statements are drafted; check editorial metadata entry at submission. +- [x] The article DOI is left to IJRC; the code uses a commit-specific URL and hash. No code DOI or preregistration is claimed. +- [ ] Confirm that the blinded manuscript, formatted supplement and masked code archive are accessible to reviewers without exposing the title page. - [ ] Re-run `scripts/reproduce_paper.py` on a clean checkout and confirm the values are unchanged. - [ ] `python -m pytest -q` green; `pytest -m slow` green. - [ ] GitHub Pages bench live, and the link works from a private window. diff --git a/papers/SUPPLEMENT.md b/papers/SUPPLEMENT.md index 086be80..e3525db 100644 --- a/papers/SUPPLEMENT.md +++ b/papers/SUPPLEMENT.md @@ -1,23 +1,23 @@ # Supplementary material -**Supplement to:** *Which Inputs Does a Network Simulation Depend On? Typed Evidence Propagation and Input-Dependence Testing, Evaluated on Cuts of a Synapse-Resolution* Drosophila *Connectome*. +**Supplement to:** *Auditing Input Dependence in Literature-Parameterised Network Simulations: A Connectome Case Study*. -FlyLab 0.6.0, commit `f95fd88a2165`, library SHA-256 `594b8d32c5e6`, map `male-cns:v1.0`. Reference numbers `[n]` refer to the manuscript's reference list. Every number here is generated by `scripts/reproduce_paper.py` from committed artifacts; none was measured in a living animal. +FlyLab 0.6.0, library SHA-256 `359096467b33`, map `male-cns:v1.0`. An internal code revision is recorded in the provenance manifest; a public revision identifier is omitted from blind-review text. Reference numbers `[n]` use the 76-entry source-paper bibliography. The standard script regenerates core model values from the current parameter library; the larger scale-study values are imported from a saved record whose 5k–50k cuts are absent here. No value was measured in a living animal. The library source-value audit remains incomplete. ## S1. Mechanism rules: the gain patch and its rationale -`flylab/pharm/mechanisms.py` is the single source of truth mapping engagement rows onto circuit gains; no assay carries its own copy, which is what makes Table T2 a truthful description of the software. Gains are dimensionless multipliers on the synaptic weight of a transmitter class (or, for `g_nav`, on node excitability); 1.0 is vehicle, and every gain is floored at 0.05 because a network with a gain of exactly zero is not informative. The 15 rules are listed in full in T2. The four shipped since v0.4 may not be changed silently: +`flylab/pharm/mechanisms.py` is the single source of truth mapping engagement rows onto circuit gains; no assay carries its own copy. Table T2 therefore describes the implemented rules. Gains are dimensionless multipliers on the synaptic weight of a transmitter class (or, for `g_nav`, on node excitability); 1.0 is vehicle, and every gain is floored at 0.05 because a network with a gain of exactly zero is not informative. The 15 rules are listed in full in T2. The four core rules are: -* **Insect nAChR, activating:** `g_ach = max(0.05, 1 + 0.4θ − 1.6θ²)`. Rises at low engagement and falls at high engagement, encoding excitation followed by desensitisation and depolarisation block. At θ = 0.99 it is at the floor, which is why a saturating dose of a nicotinic agonist silences cholinergic transmission in the model rather than merely reducing it — and, as §3.4 of the manuscript shows, why the suppression conclusion is a property of *this* rule. +* **Insect nAChR, activating:** `g_ach = max(0.05, 1 + 0.4θ − 1.6θ²)`. Rises at low engagement and falls at high engagement, encoding excitation followed by desensitisation and depolarisation block. At θ = 0.99 it reaches the model's gain floor of 0.05, sharply reducing cholinergic transmission. As §3.4 of the manuscript shows, the suppression predicate depends on this rule. * **Insect nAChR, blocking:** `g_ach = max(0.05, 1 − θ)`. * **Insect RDL, blocking:** `g_gaba = max(0.05, 1 − θ)`. Blocking the GABA-gated chloride channel removes inhibition. * **Insect RDL, activating:** `g_gaba = max(0.05, 1 + 0.4θ)`. -Four families added in v0.5 share the shape and state a rationale each: GluCl activation uses a larger coefficient (`1 + 0.8θ`) than RDL because avermectin opening is effectively irreversible [38]; AChE inhibition raises an ambient cholinergic tone (`ach_tone = 1 + 2.0θ`) multiplying `g_ach` together with the agonist curve, reproducing the organophosphate excitation-then-block picture; Nav modulation scales node excitability globally (`g_nav = 1 + 1.5θ`), since pyrethroids and DDT act on every cell regardless of transmitter [23], [67]; octopamine agonism uses the smallest coefficient (`1 + 0.5θ`) because it is modulatory. +Four additional families use stated modelling rationales: GluCl activation uses a larger coefficient (`1 + 0.8θ`) than RDL because avermectin opening is effectively irreversible [37]; AChE inhibition raises an ambient cholinergic tone (`ach_tone = 1 + 2.0θ`) multiplying `g_ach` together with the agonist curve, reproducing the organophosphate excitation-then-block picture; Nav modulation scales node excitability globally (`g_nav = 1 + 1.5θ`), since pyrethroids and DDT act on every cell regardless of transmitter [22], [66]; octopamine agonism uses the smallest coefficient (`1 + 0.5θ`) because it is modulatory. -**The level-C generic rule.** The ablation ladder's topology-only level is given one number per compound instead of a mechanism-specific gain vector. Until v0.6.1 that number was `g_all = max(0.05, 1 - theta_max)`, applied to every transmitter alike — a rule that can only depress, so every one of its 21 × 4 entries was at most zero and the level could not express disinhibition at all. It is now `g_all = max(0.05, 1 - theta_max) if the compound's dominant mechanism lowers net synaptic drive, else 1 + theta_max`: the same generic magnitude, with one bit of sign read off the mechanism table. The historical rule: max(0.05, 1 - theta_max) on every transmitter. Structurally incapable of a positive effect, so it cannot order a library containing RDL/GluCl blockers. Reported as a FLOOR (what a depression-only connectome model achieves), never as a competitor. Both are reported, the old one as `C_topology_only_floor` in T13 and T13b. +**The level-C generic rule.** The ablation ladder gives its topology-only level one signed multiplier per compound rather than a mechanism-specific vector. The direction-aware rule is `g_all = max(0.05, 1 - theta_max) if the compound's dominant mechanism lowers net synaptic drive, else 1 + theta_max`; its magnitude matches the depression-only comparison `g_all = max(0.05, 1 - theta_max)`. The historical rule: max(0.05, 1 - theta_max) on every transmitter. Structurally incapable of a positive effect, so it cannot order a library containing RDL/GluCl blockers. Reported as a FLOOR (what a depression-only connectome model achieves), never as a competitor. Both are reported for 21 compounds at 4 concentrations in T13 and T13b. Their difference assesses sensitivity to the generic rule and is not a biological contrast. -**These coefficients and shapes are modelling choices, not measurements.** That is why the manuscript prespecifies a family of 25 alternatives (§2.4) and re-derives every conclusion under each, and why the variance analysis nominates this transformation as the assumption worth measuring first. The **transmitter sign table** the rate engine iterates is a second set of asserted coefficients and is published as T21: 5 of the seven magnitudes (glutamate -0.4, histamine -0.5, dopamine +0.2, serotonin +0.2, octopamine +0.2) carry no measurement, are members of no specification family, and are not bounded by the label-permuting null, which permutes assignments and leaves the table itself untouched. +**These coefficients and shapes are modelling choices, not measurements.** The manuscript evaluates 25 gain-rule specifications (§2.4), but the five asserted transmitter sign magnitudes (glutamate -0.4, histamine -0.5, dopamine +0.2, serotonin +0.2, octopamine +0.2; T21) and transmitter-defined gain assignment remain fixed. The label-permutation null changes which nodes receive labels; it does not vary the sign magnitudes. Therefore the family does not bound uncertainty from these assumptions. ### S1.1 The specification family @@ -30,17 +30,14 @@ Five shapes — monotone-linear, saturating, weakly biphasic, the shipped biphas | C3 | imidacloprid's mean-rate effect is not distinguishable from any structure-preserving degradation of the cut (not topology-dependent) | 25 | 0 | 0 | | C4 | fipronil's mean-rate effect is topology-dependent where imidacloprid's is not, on the same cut, shuffles and criterion | 25 | 0 | 0 | | C5 | the neighborhood buffers nicotinic receptor selectivity | 18 | 0 | 7 | -| C6 | the neighborhood amplifies Nav / AChE receptor selectivity | 19 | 6 | 0 | +| C6 | the neighborhood amplifies Nav / AChE receptor selectivity | 18 | 7 | 0 | | C7 | on the map path, bitter drive lowers MN9 under fipronil | 25 | 0 | 0 | -**What the v0.6 matrix could not have told you.** Two defects made the topology rows of the previous matrix empty rather than merely noisy, and both are fixed here. +**Specification-test scope.** The topology predicates use 100 shuffles per specification, allowing empirical probabilities to reach α = 0.050. The specification is passed explicitly into the null engine. The C1–C7 rows of T14 are predicates about model outputs under this fixed implementation and selected specification family; they are not biological validation. -1. *The shuffle count could not reject.* The topology predicates were evaluated at 6 shuffles per specification, where the smallest attainable empirical permutation probability is 1/7 = 0.143. `p ≤ 0.05` was arithmetically unattainable, so the predicate fell back on a Gaussian `|z| < 2` and C3's 25/25 was a non-rejection under a test that could not reject. The run behind this document uses 100 shuffles per specification (resolution 0.010), which supports α = 0.050 with room to spare. -2. *The specification never reached the null engine.* `robustness.mechanism_spec` rebinds `compute_gains` on the assay modules, but the permutation path resolves gains through `flylab.circuit.rate.compute_gains`, which it did not rebind. Every specification therefore fed the **default** gains to its nulls and produced identical effects, so a uniform retention rate across the family was guaranteed by construction. The dependence engine now accepts the gain patch explicitly (`gains_by_compound`), and the per-specification topology verdicts in T14 are computed under the specification they are labelled with. +The topology predicates are retained in 25/25 and 25/25 cases after Benjamini-Hochberg adjustment of the 100 structural component tests (uncorrected counts 25 and 25). This adjustment does not give a cell-level or predicate-level FDR guarantee. Among C3 cases, 17 pass the descriptive point-gap criterion and 8 remain indeterminate; the C4 split is 0 / 0. These counts are conditional on the fixed sign table, assignment rule and common shuffle stream, not independent replications. -Both topology conclusions survive the correction (25/25 and 25/25 after Benjamini-Hochberg across the 100 structural tests of the run; 25 and 25 uncorrected), but a retention rate for a conclusion shaped as a *failure to reject* is only as strong as the equivalence behind it. Of C3's 25, **17 reach equivalence within the prespecified margin and 8 are merely indeterminate**; the corresponding split for C4 is 0 / 0. Every specification's topology cell is run on the same shuffle stream, so the structural tests are positively dependent rather than independent; Benjamini-Hochberg controls the FDR under positive dependence, but the fractions describe this shuffle stream and are not 25 independent replications. - -Specifications reversing C1: linear@0.50x, linear@0.75x, linear@1.00x, linear@1.25x, linear@1.50x, saturating@0.50x, saturating@0.75x, saturating@1.00x, saturating@1.25x, saturating@1.50x. Mean SI gap over the nicotinic set under the shipped rule: -0.884 log₁₀ units; averaged over the 6 reversing specifications for which the index is defined: -0.893. **The v0.6 supplement said the monotone rules "buffer at least as much, never less". That is false.** The per-specification range runs -0.199 to -1.734, 3 of them buffer *less* than the shipped rule (linear@1.50x, saturating@1.25x, saturating@1.50x), and the aggregate is carried by the extreme of the range. C5 survives the specifications that reverse C1, but not for the reason the previous version gave. +Specifications reversing C1: linear@0.50x, linear@0.75x, linear@1.00x, linear@1.25x, linear@1.50x, saturating@0.50x, saturating@0.75x, saturating@1.00x, saturating@1.25x, saturating@1.50x. The mean selectivity-index gap in the nicotinic set under the shipped gain rule is -0.884 log₁₀ units; over 6 reversing specifications with a defined index it is -0.893. The per-specification range is -0.199 to -1.734; 3 buffer less than the shipped rule (linear@1.50x, saturating@1.25x, saturating@1.50x). An aggregate gap should not be interpreted as a uniform response across specifications. ## S2. Null-model definitions, and the conventions around them @@ -52,7 +49,7 @@ Specifications reversing C1: linear@0.50x, linear@0.75x, linear@1.00x, linear@1. * **`rewire_degree_preserving`** — double-edge swaps, roughly 10 × |E| accepted swaps. Every node keeps its in-degree, out-degree and transmitter; the wiring pattern is destroyed. * **`erdos_renyi`** — the same node and edge counts, endpoints drawn uniformly, weights drawn from the empirical weight list. Only the size of the graph and its census survive. -**Why `sign_permute` is not a transmitter-identity null.** The v0.6 supplement stated that permuting labels preserves the transmitter histogram "and therefore the global excitation/inhibition balance, preserved *exactly*". That is wrong, and the error was load-bearing: out-strength is heterogeneous across nodes, so the *weighted* balance moves. Measured (T23), the cholinergic share of total outgoing synaptic weight on the `named` cut is 0.619 on the real graph against 0.544 ± 0.027 across 1000 permutations, range 0.461–0.616, putting the real graph at the 100th percentile of its own null. Because a gain patch acts *through* transmitter identity, the mode moves the drug's target set and the sign matrix at once, and the extra variance makes it conservative. The weight-matched null holds every tracked share to within 0.002. Substituting it does not change imidacloprid's permutation probability but it shrinks the equivalence gap from 2.652 Hz to 0.236 Hz against a margin of 0.332 Hz. +**Transmitter-label null.** Plain label permutation preserves class counts but changes their weighted outgoing shares because node out-strength varies. In the `named` cut, the cholinergic outgoing-weight share is 0.619 on the real graph and 0.544 ± 0.027 across 1000 permutations (range 0.461–0.616), with the observed share at the 100th percentile of that null. It changes both the compound target set and the signed matrix. The matched variant constrains each tracked weighted share to within 0.002. The imidacloprid median point gap changes from 2.652 to 0.236 rate-model units against margin 0.332. These nulls still hold the transmitter sign magnitudes fixed. Every mode preserves the node set, the node order and the seed block, and never creates self-loops or duplicate edges. The effect is always `treated − vehicle` with **both arms on the same graph**. The reported probability is the empirical two-sided permutation *p*, `(k+1)/(n+1)`, with resolution `1/(n+1)`; on a landscape the same `n` shuffled graphs are reused by every cell (a paired design), and shuffle *i* depends only on `(seed, i)`, so each cell's numbers are identical to running its profile alone. @@ -60,43 +57,43 @@ Every mode preserves the node set, the node order and the seed block, and never ### S2.1 The three-way verdict and the equivalence margin -Failing to reject a null is not evidence of equivalence, and none of the software's return values says it is. Each mode gets exactly one of `distinguishable`, `equivalent_within_tolerance` or `indeterminate`, against a margin δ prespecified as 5% of the **vehicle readout** of the real graph. The vehicle readout is the right scale because it is in the units of the readout, does not depend on the drug, on the null ensemble or on the number of shuffles, and is the same for every mode of a cell, so an equivalence claim made against it is comparable across the ladder. It is one tenth of the 50 % relative change the selectivity module treats as a circuit response. A cell whose baseline is undefined or zero can form no margin; its non-rejected modes are then `indeterminate`, which is the correct answer rather than a missing one. +Each graph mode receives `distinguishable`, `equivalent_within_tolerance` or `indeterminate` under a margin δ set at 5% of the real graph’s vehicle readout. `equivalent_within_tolerance` requires non-rejection plus an observed point gap below δ; it is a descriptive rule, not a formal equivalence test based on a confidence interval. A zero or undefined baseline yields no usable margin and non-rejected modes are indeterminate. -**The confirmatory profile in full.** For imidacloprid at 1 µM on `named`, at 1000 permutations against a margin of 0.332 Hz, the per-mode probabilities and verdicts are: weight permutation *p* = 0.586, equivalent_within_tolerance; degree-preserving rewiring *p* = 0.472, equivalent_within_tolerance; the weight-matched transmitter null *p* = 0.389, equivalent_within_tolerance; the joint label-permutation null *p* = 0.275, indeterminate; and the Erdős–Rényi control *p* = 9.99e-04, distinguishable. The resulting necessary level is `degree_sequence` with verdict equivalent_within_tolerance. +**The confirmatory profile in full.** For imidacloprid at 1 µM on `named`, at 1000 permutations against a margin of 0.332 rate-model units, the per-mode probabilities and verdicts are: weight permutation *p* = 0.586, equivalent_within_tolerance; degree-preserving rewiring *p* = 0.472, equivalent_within_tolerance; the weight-matched transmitter null *p* = 0.389, equivalent_within_tolerance; the joint label-permutation null *p* = 0.275, indeterminate; and the Erdős–Rényi control *p* = 9.99e-04, distinguishable. The resulting necessary level is `degree_sequence` with verdict equivalent_within_tolerance. -The **necessary information level** is the weakest rung the test could not distinguish from the real cut, reported with its verdict. Ranks 2 and 3 are not strictly nested — `weight_permute` keeps transmitter identity and loses the weight-topology pairing, the transmitter nulls the reverse — so a richer model can be distinguishable where a weaker one is not. Such cells are **non-monotone** and are flagged: 11 of 84 on the `named` cut (chlorpyrifos_oxon 10 nM, dieldrin 10 nM, dieldrin 100 nM, dieldrin 1 µM, fipronil 10 nM, fipronil 100 nM, ivermectin 1 µM, ivermectin 10 µM, picrotoxin 100 nM, picrotoxin 1 µM, picrotoxin 10 µM) and 0 of 84 on `taste_motor`. Their level must be read as "the cheapest graph model this test cannot tell apart from the real one", never as "everything above it is indistinguishable too". +The software field **necessary information level** reports the lowest-ranked passing degradation and its verdict; its name does not assert biological necessity. Ranks 2 and 3 are not strictly nested — `weight_permute` keeps transmitter identity and loses the weight-topology pairing, the transmitter nulls the reverse — so a richer model can be distinguishable where a weaker one is not. Such cells are **non-monotone** and are flagged: 0 of 84 on the `named` cut (none) and 1 of 84 on `taste_motor`. The reported level does not imply that richer degradations also pass. ### S2.2 Effect floor, multiplicity, and confirmatory versus exploratory -**Effect floor.** An absolute floor of 1e-6 Hz against a ~6.6 Hz baseline admits effects at numerical-noise level, so the effective floor is `max(absolute, 1 % × |vehicle readout|)`. Landscapes report class counts under both. On the `named` cut the relative floor moves 18 cells into `no-effect` (acetylcholine 10 nM (0.000761 Hz), acetylcholine 100 nM (0.0118 Hz), ddt 10 nM (0.00309 Hz), ddt 100 nM (0.0309 Hz), gaba 100 nM (-2.74e-05 Hz), gaba 1 µM (-0.000864 Hz), gaba 10 µM (-0.025 Hz), ivermectin 10 nM (-0.0156 Hz), ivermectin 100 nM (-0.0587 Hz), picrotoxin 10 nM (0.0076 Hz), spinosad 10 nM (0.00601 Hz), spinosad 100 nM (0.0549 Hz), sulfoxaflor 10 nM (0.000202 Hz), sulfoxaflor 100 nM (0.00202 Hz), sulfoxaflor 1 µM (0.0196 Hz), thiamethoxam 10 nM (0.000303 Hz), thiamethoxam 100 nM (0.00302 Hz), thiamethoxam 1 µM (0.0289 Hz)), changing the counts from 16 / 55 / 13 to 11 / 42 / 31. +**Effect floor.** The absolute floor of 1e-6 rate-model units is below relevant baseline magnitudes, so the effective floor is `max(absolute, 1 % × |vehicle readout|)`. Landscapes report both. On `named`, the relative floor moves 15 cells to `no-effect` (acetylcholine 10 nM (0.000761 rate-model units), acetylcholine 100 nM (0.0118 rate-model units), ddt 10 nM (0.00309 rate-model units), ddt 100 nM (0.0309 rate-model units), gaba 100 nM (-2.74e-05 rate-model units), gaba 1 µM (-0.000864 rate-model units), gaba 10 µM (-0.025 rate-model units), ivermectin 10 nM (-0.0125 rate-model units), ivermectin 100 nM (-0.0312 rate-model units), ivermectin 1 µM (-0.0367 rate-model units), ivermectin 10 µM (-0.0373 rate-model units), picrotoxin 10 nM (0.0076 rate-model units), thiamethoxam 10 nM (0.000303 rate-model units), thiamethoxam 100 nM (0.00302 rate-model units), thiamethoxam 1 µM (0.0289 rate-model units)). Counts under the absolute floor alone are 0 / 59 / 25; after the relative floor they are 11 / 33 / 40 before multiplicity adjustment and 0 / 44 / 40 afterward. -**Multiplicity.** A landscape is a screen. Benjamini-Hochberg is applied across the structural tests of the run (168 on `named`, 168 on `taste_motor`), the class is decided on the adjusted probabilities, and both counts are reported. Because an empirical permutation probability cannot fall below `1/(n+1)`, the result also reports how many tests must sit at that floor before anything can be rejected, so that a corrected count of zero is never mistaken for a finding (`fdr_can_reject` = yes here). +**Multiplicity.** The landscape is exploratory. Benjamini-Hochberg adjusts its declared structural component-test probabilities (168 on `named`, 168 on `taste_motor`), and classes are recomputed from those adjusted values. This does not imply an FDR guarantee for the derived cell labels. The report also states the empirical permutation resolution and whether any adjusted component test can cross α at that resolution (`fdr_can_reject` = yes). -**Confirmatory versus exploratory.** The imidacloprid and fipronil profiles at *n* ≥ 1000 are prespecified planned comparisons and carry no correction; everything else is exploratory and says so in its warnings. Fipronil at 1 µM is the clean illustration: confirmatory *p* = 0.006, but inside the exploratory landscape its adjusted probability is 0.050 and the cell does not survive correction (composition-dominated corrected against topology-dependent raw). +**Planned versus exploratory comparisons.** The imidacloprid and fipronil profiles at *n* ≥ 1000 are planned comparisons reported without multiplicity adjustment. Landscape screening is exploratory. For fipronil at 1 µM, the planned component-test *p* is 0.006; in the landscape the adjusted component-test value is 0.062 and the derived class is composition-dominated versus topology-dependent before adjustment. ### S2.3 Does the instrument work? Ground-truth recovery and power -Without this, a negative from the ladder is indistinguishable from an under-powered test. A synthetic cut of 250 nodes and about 1500 edges is built with log-normal weights and a transmitter composition close to the real cut's, and a directed cycle of cholinergic nodes passing through the driven seeds is planted in it at a known strength. The cycle is recurrent cholinergic gain, so a patch that collapses `g_ach` removes an amplification that exists only while the cycle is intact: the drug *contrast*, not merely the rate, depends on the wiring. Strength 0 plants nothing and is the negative control. Nothing here touches the connectome or the compound library. +The planted-cycle synthetic test examines whether the procedure recovers one known wiring-dependent contrast. A synthetic cut of 250 nodes and about 1500 edges is built with log-normal weights and a transmitter composition close to the real cut’s. A directed cholinergic cycle through the driven seeds is planted at known strength; strength 0 is the negative control. This test does not estimate power for arbitrary circuit effects or validate the pharmacological parameter library. -the ladder recovered 2 of 3 planted topology-dependent effects at n = 200 and returned no false positive on the unplanted control. Over the power grid (6 independently generated graphs per cell, planted strengths 0, 0.25, 0.5, 1, permutation counts 50, 200, 1000) the detection rate by strength is strength 0.25: 0.00-0.17; strength 0.5: 0.67-0.83; strength 1: 1.00-1.00 and by permutation count n=50: 0.61, n=200: 0.61, n=1000: 0.67; the empirical false-positive rate on the unplanted control is 0 of 18 (0.000). Effect size dominates: raising the permutation count buys far less than raising the planted strength. Full grid: T22, Fig. F16. +the ladder recovered 2 of 3 planted topology-dependent effects at n = 200 and returned no false positive on the unplanted control. In a finite grid of 6 independently generated graphs per cell, planted strengths 0, 0.25, 0.5, 1 and permutation counts 50, 200, 1000, detection by strength is strength 0.25: 0.00-0.17; strength 0.5: 0.67-0.83; strength 1: 1.00-1.00 and by count n=50: 0.61, n=200: 0.61, n=1000: 0.67. The unplanted control has 0 detections in 18 runs (0.000). These frequencies are specific to this planted mechanism and grid. **Permutation counts in this document.** 1000 per mode for the `named` profiles (resolution 0.0010), 300 for the `taste_motor` path arm, 1000 for each of the 84 `named` landscape cells and 300 for each of the 84 `taste_motor` cells. The sweep (T11) shows the verdict at α = 0.05 settling from *n* ≈ 50–400, and a mid-range *p* reaching ±0.02 of its final value between *n* = 25 and 400. ### S2.4 Where does the verdict settle with the size of the cut? -The verdict reversed between the two committed cuts, and two cuts are not a trend. `flylab.analysis.scale.dependence_vs_scale` runs the *same* dependence profile (`dependence_profile`, unchanged) across a ladder of nested cuts of 1k, 5k, 10k, 25k, 50k cells grown from the taste-motor seed set by one fixed rule, so a verdict can be compared across size without the construction changing underneath it. The two committed cuts are built by a different recipe, so the clean scale axis is the `scale_*` rungs alone; a sequence that interleaves the committed cuts mixes a size effect with a construction effect, and `verdict_stability(recipe="scale_ladder")` reports each separately. A verdict counts as *settled* only if it is the same on the two largest cuts run with no change above that point — one rung of agreement is n = 1 and is not stability. +The descriptive dependence labels differ between the two committed cuts. A saved scale-study record contains a nested-cut ladder of 1k, 5k, 10k, 25k, 50k cells grown by a fixed ranked-BFS recipe. The 5k–50k input cuts are absent from this package, so their recorded results cannot be independently rebuilt here. The two committed cuts use another construction rule, and contrasts between them confound size, density, composition and extraction. Cost is the constraint, not correctness. From the per-shuffle and per-edge constants measured on this machine (T26), a profile at *n* = 1000 on the 50k rung is 21.1 hours on one core and the landscape over the specification family is 7 days per specification; a whole-CNS profile at the weight-5 floor is 59 hours and the dense LIF matrix stops fitting in memory well before it. The permutation count must therefore fall with the cut, and `permutation_budget` refuses to go below the resolution floor `1/α − 1`, so every rung records the *n* it was run at. -The ladder's smallest rung, `scale_1k`, is already the decisive contrast with the two cuts of T19: 1000 nodes, 22 857 edges, mean degree 22.9, in-star no — *fewer* nodes than `named` and a densely recurrent cut nothing like it. +The saved ladder's smallest rung, `scale_1k`, provides a descriptive comparison with the committed cuts of T19: 1000 nodes, 22 857 edges, mean degree 22.9, in-star no — fewer nodes than `named`, but a denser cut under a different construction rule. -**Result** (T27; the study record is read from `scale_study.json` beside the paper artifacts, or from `data/derived/` when it is promoted there). Over the 7 cuts run (scale_1k, named, taste_motor, scale_5k, scale_10k, scale_25k, scale_50k), the composition-dominated verdict survives on 1 — named — so the inversion does not track the number of cells: `scale_1k` has 1000 nodes against `named`'s 1126 and a mean degree of 22.9 against 1.21, and it is already topology-dependent. Per-rung sequences: scale_1k (1000 nodes, mean degree 22.9): topology-dependent; named (1126 nodes, mean degree 1.2): composition-dominated; taste_motor (1841 nodes, mean degree 10.4): topology-dependent; scale_5k (5000 nodes, mean degree 56.0): topology-dependent; scale_10k (10000 nodes, mean degree 62.2): topology-dependent; scale_25k (25000 nodes, mean degree 54.6): topology-dependent; scale_50k (50000 nodes, mean degree 44.3): topology-dependent. Fipronil is topology-dependent on every rung. In summary: imidacloprid: the verdict is 'topology-dependent' on the largest 5 of 7 cuts tested, first reached at 'taste_motor'. fipronil: the verdict is 'topology-dependent' on the largest 7 of 7 cuts tested, first reached at 'scale_1k'. On the clean `scale_*` ladder the verdicts settle: yes. At scale_5k, scale_10k the necessary information level is the real network itself, i.e. the prediction is distinguishable from *every* rung of the ladder. Imidacloprid's relative effect stays within 0.52-0.93 of the vehicle readout across the ladder, so the verdict changes while the quantity being measured does not. +**Imported result** (T27, from saved `scale_study.json`). Across 7 recorded cuts (scale_1k, named, taste_motor, scale_5k, scale_10k, scale_25k, scale_50k), the composition-dominated label occurs on 1 (named). Per-rung imidacloprid labels are scale_1k (1000 nodes, mean degree 22.9): topology-dependent; named (1126 nodes, mean degree 1.2): composition-dominated; taste_motor (1841 nodes, mean degree 10.4): topology-dependent; scale_5k (5000 nodes, mean degree 56.0): topology-dependent; scale_10k (10000 nodes, mean degree 62.2): topology-dependent; scale_25k (25000 nodes, mean degree 54.6): topology-dependent; scale_50k (50000 nodes, mean degree 44.3): topology-dependent; the record labels fipronil topology-dependent on each rung. The record summaries are imidacloprid: the verdict is 'topology-dependent' on the largest 5 of 7 cuts tested, first reached at 'taste_motor'. fipronil: the verdict is 'topology-dependent' on the largest 7 of 7 cuts tested, first reached at 'scale_1k'. and the inversion does not track the number of cells: `scale_1k` has 1000 nodes against `named`'s 1126 and a mean degree of 22.9 against 1.21, and it is already topology-dependent. It reports scale_5k, scale_10k cuts where imidacloprid is distinguishable from all three ranked edge/wiring degradations (`weight_permute`, `rewire_degree_preserving`, `erdos_renyi`). The orthogonal transmitter-label checks are separate: plain `sign_permute` is indeterminate at `scale_10k`. The imidacloprid relative effect lies within 0.52-0.93 of the vehicle readout, and the clean-ladder summary is yes. These are conditional observations from the saved run; the absent larger input cuts prevent independent regeneration of all rungs. -**Two caveats, both reported rather than absorbed.** The permutation budget falls with the edge count (20-1000 draws, resolution 0.0010-0.0476): at scale_10k, scale_25k, scale_50k the permutation budget falls to n = 20, where the smallest attainable probability is 0.048 against alpha = 0.05: a rejection there is the smallest the test can express, so those rungs confirm a verdict already settled lower down rather than establishing it. And `named` and `taste_motor` are hops-limited neighbourhoods while the `scale_*` rungs are nested ranked-BFS cuts, so a difference *between* the two families confounds scale with construction; differences *within* the `scale_*` rungs do not. +**Scale-study limits.** Permutation counts decrease as edge count grows (20-1000 draws; resolutions 0.0010-0.0476), and at scale_10k, scale_25k, scale_50k the permutation budget falls to n = 20, where the smallest attainable probability is 0.048 against alpha = 0.05: a rejection there is the smallest the test can express. The `named` and `taste_motor` cuts are hops-limited whereas the `scale_*` ladder is ranked-BFS. No comparison here identifies recurrence, density or scale as an isolated cause of the class change. ## S3. Runtimes and their calibration -**Rate runtime.** `flylab/circuit/rate.py` holds a dense signed weight matrix `W[j,i] = sign(nt_i) × synapses(i→j)`, with the signs of Table T21. **That matrix is then row-normalised**: `W_signed = W / max(Σ_i |W[j,i]|, 1)`, one denominator per postsynaptic cell. A gain patch afterwards scales the columns by the presynaptic transmitter's gain, so `W_eff = W_signed · diag(g)` and **each cell's recurrent input is a composition-weighted average of its presynaptic gains, normalised by its own total absolute input**. The update is a clipped leaky iteration `r ← clip((1 − α)·r + α·g_nav·(d + W_eff·r), 0, r_max)` with α = 0.3, 80 steps, `r_max` = 300 Hz and *d* the per-node external drive in Hz. It is deterministic, and one evaluation of the 1126-node graph costs about 50 ms with single-threaded BLAS — which is what makes 1000-permutation profiles and an 84-cell landscape affordable at all. +**Rate runtime.** `flylab/circuit/rate.py` constructs `W[j,i] = sign(nt_i) × synapses(i→j)` and row-normalises it as `W_signed = W / max(Σ_i |W[j,i]|, 1)` before applying presynaptic gains: `W_eff = W_signed · diag(g)`. The clipped iteration is `r ← clip((1 − α)·r + α·g_nav·(d + W_eff·r), 0, r_max)` with α = 0.3 and 80 steps. Its numerical readout is in **rate-model units**, not calibrated firing hertz; drive and clipping parameters are set on the model scale. This deterministic runtime takes about 50 ms for one 1126-node evaluation with single-threaded BLAS. The normalisation is an **asserted modelling choice**, not a measurement, and it is consequential for RQ2: a readout in which each cell averages its presynaptic gains builds part of "composition-dominated" into itself. Three modes are implemented and swept in §3.3 (T25): @@ -106,23 +103,23 @@ The normalisation is an **asserted modelling choice**, not a measurement, and it | `none` | 1 | denom[j] = 1: raw signed synapse counts. The recurrent operator is supercritical (spectral radius ~150 on the named cut) and the r_max clip, not the wiring, bounds the rates. | | `degree` | `max(in-degree, 1)` | denom[j] = max(in-degree[j], 1): degree-corrected. Relative input strengths survive (unlike row_abs) but hubs are discounted by fan-in. | -The composition-versus-full rank correlation at 1 µM is 0.988 under `row_abs`, 0.906 under `none` and 0.666 under `degree` (-0.372 at 10 nM): the composition verdict survives removing the normalisation but not replacing it with a degree-corrected one. Levels are not comparable across modes — the unnormalised operator is supercritical on this cut, so the `r_max` clip rather than the wiring bounds the rates — only the orderings are. +The composition-versus-full rank correlation at 1 µM is 0.921 under `row_abs`, 0.935 under `none` and 0.551 under `degree` (-0.583 at 10 nM): the composition verdict survives removing the normalisation but not replacing it with a degree-corrected one. Levels are not comparable across modes — the unnormalised operator is supercritical on this cut, so the `r_max` clip rather than the wiring bounds the rates — only the orderings are. **LIF runtime.** `flylab/circuit/lif.py` is a current-based leaky integrate-and-fire network with the parameters of Shiu *et al.* [7]: τ_m = 20 ms, V_rest = −52 mV, V_th = −45 mV, V_reset = −52 mV, refractory 2.2 ms, exponential synaptic current with τ_s = 5 ms, forward Euler at dt = 0.1 ms. A presynaptic spike on edge *i*→*j* injects `w_scale × synapses × sign(nt_i) × gain(nt_i)` mV into *j*. -Two choices are modelling decisions, not literature values. *The synaptic scale is a one-time calibration scan*: `w_scale` = 0.05 mV per synapse and a drive weight of 0.5 mV per event were chosen once, by scanning, so that vehicle MN9 cells fire in a plausible band under the standard 150 Hz seed drive. No recording was used and no error was minimised against data, so this does not consume the one fitted parameter the project allows — but it is a tuned constant. *A background Poisson drive is required, and it is a choice*: excising a 1-hop neighbourhood removes every input from outside it, so every non-seed cell receives an independent Poisson drive at 0.65 × the seed rate, standing in for the amputated inputs. This is the most consequential free choice in the LIF arm; it sets the operating point and it is why the two engines report sugar-driven MN9 at 0.293 Hz (rate) against 105 Hz (LIF). +Two LIF choices are modelling decisions, not measured values. The synaptic scale `w_scale` = 0.05 mV per synapse and drive weight 0.5 mV per event were selected by a one-time scan, without fitting to recordings. Background Poisson drive at 0.65 × seed rate is supplied to non-seed cells to stand in for inputs omitted by the 1-hop cut. It sets the LIF operating point. The rate-model output is 0.293 rate-model units; the LIF output is 105 Hz under the standard 150 Hz seed drive. The values are not directly calibrated to one another. ## S4. Supporting capabilities These are implemented, tested and used, but are supporting capabilities rather than contributions of this paper. -**Exposure.** `flylab/pharm/exposure.py` maps a dose and route onto *C(t)* and hence engagement(t) in one compartment. Its constants are class placeholders chosen so that the three routes differ in the direction the literature describes, and are fitted to no *Drosophila* ADME dataset because that dataset does not exist: of the seven quantities such a model needs, `data/literature/fly_pharmacokinetics.yaml` records only metabolite identity as well established [52], [54], [55], [58]. Adult haemolymph volume, adult body mass, cuticular penetration rate and any *Drosophila* elimination half-life are `null`; what is recorded is 25 nL recoverable per adult [57], 174 ± 81 nL in third-instar larvae [56], a water-partitioning measurement [53], an ethanol entry that could not supply a rate [65], and an imidacloprid half-life of 4.5–5 h in *honeybee* [58] — the only verified insect half-life for a library compound. Two structural warnings travel with every profile: haemolymph is the lowest-radioactivity compartment for imidacloprid, so a one-compartment haemolymph model under-estimates CNS exposure [59]; and a substantial metabolic route runs through the gut microbiome, so a single-enzyme clearance term is wrong for this compound [54]. +**Exposure.** `flylab/pharm/exposure.py` maps a dose and route onto *C(t)* and hence engagement(t) in one compartment. Its constants are class placeholders chosen so that the three routes differ in the direction the literature describes; they are not fitted to a *Drosophila* ADME dataset. Of the seven quantities such a model needs, `data/literature/fly_pharmacokinetics.yaml` records scattered metabolism observations from *Drosophila*, heterologous cell culture and honeybees [53], [54], [57], but not a complete *Drosophila* exposure series. Adult haemolymph volume, adult body mass, cuticular penetration rate and any *Drosophila* elimination half-life are `null`; what is recorded is 25 nL recoverable per adult [56], 174 ± 81 nL in third-instar larvae [55], a water-partitioning measurement [52], an ethanol entry that could not supply a rate [64], and an imidacloprid half-life of 4.5–5 h in *honeybee* [57] — the only verified insect half-life for a library compound. Two structural warnings travel with every profile: a honeybee study found lower total radioactivity in haemolymph than in other sampled compartments [58], so a one-compartment haemolymph model may miss tissue distribution; the direction and size of any CNS exposure bias are unknown. A substantial metabolic route through the gut microbiome also makes a single-enzyme clearance term incomplete for this compound [53]. -**Genotype.** `flylab/pharm/genotype.py` applies published target-site resistance alleles as multiplicative potency shifts from `data/literature/resistance_alleles.yaml` [13]–[25], enforcing one rule: **a shift applies per compound, never per receptor.** Two published findings make the naive implementation wrong — *Rdl* A301S shifts GABA potency 33.9-fold but fipronil only 1.1-fold on the expressed planthopper receptor [15], and *para* M918T shifts deltamethrin 100-fold while leaving DDT at 1-fold [22]. An allele/compound pair with no sourced number is left unchanged and listed under `unshifted`, never extrapolated. One row is deliberately ugly: *Rdl* A301G in *D. simulans* carries a 20000-fold shift derived from a whole-animal resistance ratio, which drives fipronil engagement to 4.63e-04 and is almost certainly an overstatement of the target-site change [19]. The notebook says so. Panel: T9, F9. +**Genotype.** `flylab/pharm/genotype.py` applies published target-site resistance alleles as multiplicative potency shifts from `data/literature/resistance_alleles.yaml` [12]–[24], enforcing one rule: **a shift applies per compound, never per receptor.** Two published findings make the naive implementation wrong — *Rdl* A301S shifts GABA potency 33.9-fold but fipronil only 1.1-fold on the expressed planthopper receptor [14], and *para* M918T shifts deltamethrin 100-fold while leaving DDT at 1-fold [21]. An allele/compound pair with no sourced number is left unchanged and listed under `unshifted`, never extrapolated. One row is deliberately ugly: *Rdl* A301G in *D. simulans* carries a 20000-fold shift derived from a whole-animal resistance ratio, which drives fipronil engagement to 4.63e-04 and is almost certainly an overstatement of the target-site change [18]. The notebook says so. Panel: T9, F9. -**Mixtures.** `flylab/pharm/mixtures.py` implements Bliss independence [45] and Loewe additivity [48], [49] as explicit null models, following the response-surface framing of Greco *et al.* [46], and selects between them per pair: Loewe where the components share a receptor key, Bliss where their targets are disjoint. The three-case comparison against the honeybee dataset [51] is a qualitative concordance check and not validation. Table T5. +**Mixtures.** `flylab/pharm/mixtures.py` implements Bliss independence [44] and Loewe additivity [47], [48] as explicit null models, following the response-surface framing of Greco *et al.* [45], and selects between them per pair: Loewe where the components share a receptor key, Bliss where their targets are disjoint. The three-case comparison against the honeybee dataset [50] is a qualitative concordance check and not validation. Table T5. -**Expression weighting.** `flylab/pharm/expression.py` relaxes the assumption that every cell expresses the receptor, using class-level statements from single-cell atlases [39]–[44] mapped by hand onto MaleCNS superclasses, and interpolates each node's gain between 1.0 and the full scalar gain. It is a **sensitivity layer and never the primary model**: none of the cited atlases uses MaleCNS superclass labels, so every mapping row is tagged `cross_atlas_inference`; coverage is 0.205 of cells; and receptor expression in identified adult motor neurons is a confirmed gap, so MN9 carries the `unknown` default. Weighting shifts MN9 by -2.622 Hz at 1 µM imidacloprid, which we report as the size of the uncertainty rather than as a correction. +**Expression weighting.** `flylab/pharm/expression.py` maps class-level atlas statements [38]–[43] to MaleCNS superclasses by hand and interpolates gains. This is a sensitivity layer: the atlas labels do not match MaleCNS superclasses, every mapping is tagged `cross_atlas_inference`, 0.205 of graph-node × receptor-key pairs across the two committed cuts have a coarse class-level annotation; this is annotation availability, not the fraction of cells measured to express a receptor. Identified MN9 motor-neuron expression remains unknown. The rate-model MN9 output shifts by -2.622 rate-model units at 1 µM imidacloprid. This is a model sensitivity, not an experimental correction. **Model dose-response.** `flylab/assays/ensemble.py` builds the model's own concentration-response curve, fits a four-parameter Hill function by Nelder–Mead in log-concentration space and bootstraps it. For imidacloprid on the `named` cut the circuit midpoint is 34.7 nM with a bootstrap 95 % CI of 35.2 nM to 71.5 nM (r² = 0.9999). The interval does not bracket the point estimate, and that is reported rather than hidden: the bootstrap resamples replicates within each concentration, the lower plateau of the ladder is short, and the midpoint is therefore poorly identified. This is in any case the **model's** circuit midpoint, not a receptor or an animal IC50. The one-at-a-time tornado at 1 µM finds exactly zero span for ec50, hill_n — the engagement is saturated — which the global analysis (§3.5) confirms and quantifies. @@ -131,16 +128,16 @@ These are implemented, tested and used, but are supporting capabilities rather t The following 7 hypotheses are written before any experiment and labelled `software_prediction`; `live_result` is `null` in every row and stays null until a real table is imported by hand. They are **prospective, not pre-registered**: a mutable repository is not a registry, and the wording will change only if a tagged release is externally archived with a timestamp. Full table: T3. * **H1** (receptor engagement). At 1 µM, imidacloprid insect nicotinic engagement ≥ 0.9 while vertebrate α4β2 stays ≤ 0.2. Predicted gap 0.854. Checkable against published expressed-receptor work rather than against flies. -* **H2** (PER, `protocols/per.md`). The same dose lowers cholinergic drive in the MN9/DNp01 neighbourhood: predicted change in network mean rate -6.145 Hz. **Specification-dependence caveat.** The *sign* of this prediction is a modelling choice, not a model output: §3.4 shows the same compound at the same engagement takes the network from 6.631 Hz to 0.458 Hz under the shipped biphasic rule and from 6.631 Hz to 10.1 Hz under a monotone rule of the same admissible family, and only 15 of 25 specifications retain the suppression direction. An experimenter should treat H2 as a signed interval over the family rather than as a point prediction, and a live result in either direction falsifies a specification rather than the integration. The control rate this needs could not be sourced; the nearest substitute is 40 % responders at 1 h food deprivation rising to 68 % at 2 h in 463 wild-type flies [61], which also shows that deprivation time is a larger nuisance variable than most effects FlyLab would try to detect. -* **H3** (climbing, `protocols/climbing.md`). Fipronil at 1 µM disinhibits the neighbourhood: predicted change 0.896 Hz, opposite in sign to H2. **The same caveat applies transitively**: H3 is defined relative to H2, so its contrast inherits H2's specification dependence even though the fipronil arm itself is retained by 25 of 25 specifications. Negative geotaxis is the best-served endpoint [60], [64], [66]. -* **H4** (negative control). Diazepam engages vertebrate GABA-A and not insect RDL in this library, so it moves no fly readout: predicted effect 0.000. A non-zero live effect would falsify either the library row [17] or the assumption that the compound reaches the CNS unmetabolised. -* **H5** (within-animal control). Bitter input vetoes sugar-driven MN9 on the map-extracted path: predicted change -0.296 Hz relative to sugar alone, matching the published direction [7], [63]. -* **H6**. Fipronil relieves the bitter veto on the map path: predicted rise in the veto ratio 0.413. The dependence analysis shows this readout is not distinguishable from its shuffled controls, so a live confirmation would support the pharmacology and **not** the wiring. -* **H7** (selectivity negative control). Picrotoxin perturbs the fly neighbourhood while its receptor selectivity index is ≈ 0: predicted change 0.340 Hz. Any insect-selectivity claim the tool makes must fail for picrotoxin, and picrotoxin's lack of insect-over-vertebrate selectivity is independently attested [31]. +* **H2** (PER, `protocols/per.md`). At the stated dose, six paired model replicates give a mean change of -6.145 rate-model units in the MN9/DNp01 neighbourhood (T3). In a separate deterministic profile, the shipped rule gives 6.631 to 0.458 rate-model units, while a monotone member gives 6.631 to 10.1; 15 of 25 specifications retain suppression. The replicate mean and deterministic endpoint difference use different calculations. These are model outputs for prospective testing, not validated fly effects. The nearest sourced behavioural control concerns food-deprivation time [60], which differs from the model readout. +* **H3** (climbing, `protocols/climbing.md`). The fipronil arm predicts 0.896 rate-model units in the neighbourhood. Its comparison with H2 inherits H2’s specification dependence; 25 of 25 gain-rule specifications retain the fipronil predicate. Negative geotaxis provides a prospective assay [59], [63], [65], not current validation. +* **H4** (negative control). Diazepam engages vertebrate GABA-A and not insect RDL in this library, so it moves no fly readout: predicted effect 0.000. A nonzero biological effect would challenge the combined model assumptions; identifying its cause would require further experiments. +* **H5** (within-animal control). The extracted sensory path predicts a change of -0.296 rate-model units when bitter input accompanies sugar drive. Its direction is qualitatively concordant with the published sensory-veto result [7], [62]; the present cut and parameter choices are not an independent validation of that experiment. +* **H6**. Fipronil relieves the bitter veto on the map path: predicted rise in the veto ratio 0.413. The dependence analysis shows this readout is not distinguishable from its shuffled controls. A matching biological result would provide concordance with the predicted direction, but these null tests cannot determine whether the wiring contributes. +* **H7** (selectivity negative control). The model predicts a fly-neighbourhood change of 0.400 rate-model units for picrotoxin while its selectivity index is near 0. This checks the software’s selectivity interpretation; receptor-level cross-species data [30] do not validate the circuit output. Rows where the model's own output disagrees with the prospective direction: none. -**On sample sizes.** Model-internal intervals (6 replicates of library jitter, drive jitter and the RNG seed) contain no biological variance, so the standardised effect sizes they imply are absurd. T3 therefore caps *d* at 1.0 before computing *n*, which yields 23 per group for every row. **This is an illustrative planning minimum, not a power calculation**: it means "what a large effect would need", it assumes independent flies where vial-based protocols violate that, and it must not be used to size a study. +**On sample sizes.** Model-internal intervals (6 replicates of library jitter, drive jitter and the RNG seed) contain no biological variance, so effect sizes derived from them cannot support a biological power calculation. T3 therefore caps *d* at 1.0 before computing *n*, which yields 23 per group for every row. **This is an illustrative calculation, not a study power analysis**: it assumes a large standardised effect and independent flies, whereas vial-based protocols may violate independence; it must not be used to size a study. ## S6. Evidence typing: the full rules @@ -154,9 +151,9 @@ The permitted transformation is a function of two facts — the **parameter type A proxy computes the same Hill expression as its non-proxy sibling but reports itself as a transfer and carries a provenance warning naming what was transferred — the species, the preparation or the chemical class. Asking for a binding occupancy from an EC50, or for any numeric engagement from a placeholder row, raises `EvidenceTypeError`; the library validator rejects a placeholder that carries a value and a non-placeholder that does not. The library currently reports none validation problems. -**The soundness invariant.** *No numeric engagement reaching a circuit readout may originate from a row whose (parameter type, evidence distance) pair does not admit one.* `tests/test_pharm_evidence.py` checks it exhaustively rather than at a call site: over every row of the shipped library and every entry point that can produce an engagement, and with a companion check that dropping the inadmissible rows changes no gain vector. Mutating the rule table makes the test fail with a message naming the row and the pair. This, rather than the enum, is what the manuscript claims as a computing contribution. +**The admission invariant.** *No numeric engagement reaching an enumerated circuit-readout path may originate from a row whose (parameter type, evidence distance) pair does not admit one.* `tests/test_pharm_evidence.py` checks every shipped row and tested entry point, verifies that dropping inadmissible rows leaves the tested gain vectors unchanged, and detects a rule-table mutation. This tests software admission logic; it does not verify that every admitted number matches its cited assay. -**Census.** 102 rows over 21 compounds and 15 receptor keys (Table T10, Figure F11). Parameter types: 34 EC50, 10 IC50, 2 Kd, 56 unknown. Evidence distances: 10 E0, 15 E1, 10 E2, 11 E3, 56 E4. Engagement models: 1 `binding_occupancy`, 1 `binding_engagement_proxy`, 9 `functional_engagement`, 35 `functional_engagement_proxy`, 56 `not_modelled`. Rows entitled to a binding-occupancy model: imidacloprid at insect_nAChR_native_dmel (Kd 2e-09 M) [77]. Rows demoted to a binding proxy by distance: imidacloprid at insect_nAChR_beta1 (Kd 8.3e-11 M). +**Census.** 102 rows over 21 compounds and 15 receptor keys (Table T10, Figure F11). Parameter types: 29 EC50, 8 IC50, 2 Kd, 63 unknown. Evidence distances: 7 E0, 15 E1, 8 E2, 9 E3, 63 E4. Engagement models: 1 `binding_occupancy`, 1 `binding_engagement_proxy`, 6 `functional_engagement`, 31 `functional_engagement_proxy`, 63 `not_modelled`. Rows entitled to a binding-occupancy model: imidacloprid at insect_nAChR_native_dmel (Kd 2e-09 M) [76]. Rows demoted to a binding proxy by distance: imidacloprid at insect_nAChR_beta1 (Kd 8.3e-11 M). ## S7. Reproduction details @@ -177,7 +174,7 @@ A proxy computes the same Hill expression as its non-proxy sibling but reports i | dose-response bootstrap | 200 resamples | 40 | | prediction replicates | 6 | 2 | -The middle column is read from `results.json`, so it describes the run that rendered this document rather than an aspiration. `--fast` changes statistical effort only, never the model: point estimates that do not depend on replicate or permutation count are identical, and a `--fast` *p* is coarser. Two tests guard the committed record: it must not come from a `--fast` run, and — new in this revision — **every statistical parameter it records must equal the shipped default**, because the v0.6 record was generated at a shuffle count the code no longer used and nothing detected it. +The middle column is read from `results.json`, so it describes the run that rendered this document rather than an aspiration. `--fast` changes statistical effort only, never the model: point estimates that do not depend on replicate or permutation count are identical, and a `--fast` *p* is coarser. Two tests guard the committed record: it must not come from a `--fast` run, and every recorded statistical parameter must match the shipped default. Two of the analyses are started in worker processes when the run begins and joined at their own step, because they dominate the wall clock and depend on nothing the other steps produce. `--jobs` controls that and the permutation workers; results are identical at any value. @@ -189,11 +186,11 @@ Every assay returns a notebook (schema 0.3) whose `warnings` list ships with the These are summarised in §4 and given here at length. -**(i) The transmitter sign table is mostly asserted.** Acetylcholine at +1.0 is the normalisation every other entry is scaled against, and GABA at −1.0 is the convention that inhibition is as strong per synapse as excitation. The remaining five — glutamate -0.4, histamine -0.5, dopamine +0.2, serotonin +0.2, octopamine +0.2 — are asserted magnitudes with no measurement behind them (T21). Glutamate in particular is signed inhibitory at 40 % of the stated strength, which is a hedge against the cells whose glutamatergic output is excitatory rather than a measurement of anything; any such cell is mis-signed in this model. None of these five is a member of the specification family, and the transmitter nulls do not bound them, because they permute *assignments* and leave the table untouched. Separately, the ablation ladder's level B inherits a whole-CNS excitation index that scores glutamate as partly excitatory while level D signs it inhibitory; the two levels of one ladder therefore disagree about a transmitter sign, and reconciling them moves the composition-versus-full rank correlation by at most 0.057. +**(i) The transmitter sign table is mostly asserted.** Acetylcholine at +1.0 is the normalisation every other entry is scaled against, and GABA at −1.0 is the convention that inhibition is as strong per synapse as excitation. The remaining five — glutamate -0.4, histamine -0.5, dopamine +0.2, serotonin +0.2, octopamine +0.2 — are asserted magnitudes with no measurement behind them (T21). Glutamate in particular is signed inhibitory at 40 % of the stated strength, which is a hedge against the cells whose glutamatergic output is excitatory rather than a measurement of anything; any such cell is mis-signed in this model. None of these five is a member of the specification family, and the transmitter nulls do not bound them, because they permute *assignments* and leave the table untouched. Separately, the ablation ladder's level B inherits a whole-CNS excitation index that scores glutamate as partly excitatory while level D signs it inhibitory; the two levels of one ladder therefore disagree about a transmitter sign, and reconciling them moves the composition-versus-full rank correlation by at most 0.086. **(ii) The row normalisation is load-bearing** (§S3, §3.3). -**(iii) Receptor effects are applied across transmitter-defined synapses.** A gain on `g_ach` scales every cholinergic synapse, whether or not that synapse expresses the receptor the compound binds. This is the assumption the expression layer relaxes, and it is reported as a bound rather than an improvement: coverage is 0.205 of cells, every mapping row is a cross-atlas inference, and receptor expression in identified adult motor neurons — MN9's class — is a confirmed gap in the atlas literature, so the paper's principal readout carries the default weight. Weighting shifts MN9 by -2.622 Hz at 1 µM imidacloprid. +**(iii) Receptor effects are applied across transmitter-defined synapses.** A gain on `g_ach` scales every cholinergic synapse, whether or not that synapse expresses the receptor the compound binds. This is the assumption the expression layer relaxes, and it is reported as a sensitivity analysis: 0.205 of graph-node × receptor-key pairs across the two committed cuts have a coarse class-level annotation; this is not a fraction of cells measured to express a receptor. Every mapping row is a cross-atlas inference. We reprocessed the four adult VNC count matrices in GEO GSE141807 into gene-level and target-family detection fractions after the published QC filters; the summary is VNC-wide and contains no cell-class assignments. Three replicate cell counts match the paper; one retains two extra cells under the published thresholds, an unresolved discrepancy recorded with the analysis. Single-cell dropout means a zero count is not evidence of absence. These data do not identify MN9, a proboscis motor neuron whose arbors are in the brain's subesophageal zone, so the principal readout still carries the default weight. Weighting shifts the rate-model MN9 readout by -2.622 rate-model units at 1 µM imidacloprid. **(iv) Hops-limited cuts.** A 1-hop neighbourhood with a weight floor is a fraction of MN9's real input; missing inhibition can flip the sign of any readout, and the disinhibition readout is bounded by the GABA share that survived the cut. Growing the cut is deliberately constrained: past about 1 MB of JSON the weight floor is raised rather than hops added, because a 2-hop cut of a motor neuron in a 25-million-edge CNS is most of the animal. §3.2 is the quantitative demonstration of how much the choice of cut matters — the dependence verdict itself reverses between the two committed cuts. @@ -205,12 +202,12 @@ These are summarised in §4 and given here at length. These are the domain-side results the method is demonstrated on. They are supporting material: none of them is a contribution of this paper, and §3 evaluates the instruments rather than these numbers. -**The evidence census in full.** Of 102 rows over 21 compounds and 15 receptor keys, 34 carry an EC50, 10 an IC50 and 2 a binding constant. By evidence distance the split is 10 / 15 / 10 / 11 / 56 rows at E0–E4. At the 1 µM working concentration the two headline compounds engage their insect targets at 0.991 and 0.985 against 0.091 and 0.476 at the vertebrate counterparts (500-fold and 37-fold; 2.70 log₁₀ units of window for the first). Placeholder rows contribute nothing: the negative-control row reports "not modelled (N/A)". +**The evidence census in full.** Of 102 rows over 21 compounds and 15 receptor keys, 29 carry an EC50, 8 an IC50 and 2 a binding constant. By evidence distance the split is 7 / 15 / 8 / 9 / 63 rows at E0–E4. At the 1 µM working concentration the two headline compounds engage their insect targets at 0.991 and 0.985 against 0.091 and 0.476 at the vertebrate counterparts (500-fold and 37-fold; 2.70 log₁₀ units of window for the first). Placeholder rows contribute nothing: the negative-control row reports "not modelled (N/A)". -**Sensory-veto readout.** Driving the 34 sweet GRN body IDs at 150 Hz and reading MN9 reproduces the published direction of a bitter veto [7], [63] on both runtimes (0.000 rate, 0.000 LIF in vehicle; 0.413 and 0.159 under fipronil). The two runtimes agree on **direction only** — sugar-driven MN9 is 0.293 Hz on the rate engine against 105 Hz on the LIF engine, a difference following from the LIF background drive (§S3) — and we do not present that agreement as cross-model validation. +**Sensory-veto readout.** Driving 34 sweet GRN IDs at 150 Hz produces the published *direction* of bitter veto [7], [62] in both runtimes (vehicle ratios 0.000 rate model and 0.000 LIF; fipronil ratios 0.413 and 0.159). This is qualitative concordance only. Sugar-driven MN9 is 0.293 rate-model units versus 105 Hz in the LIF network; these numbers have different calibration status and are not cross-model validation. -**A properly powered negative.** At 300 permutations the veto ratio on the extracted sensory path is not distinguishable from any shuffled control (max |z| = 0.220), equivalent within tolerance for none and indeterminate for erdos_renyi, rewire_degree_preserving, sign_permute, sign_permute_weight_matched, weight_permute. Because §3.1 characterises the instrument's power, this is reportable as a negative rather than as an absence of evidence. The comparator compound has no veto ratio at this dose at all — the readout is *undefined*, not zero, and is reported as undefined. +**Path-null result.** At 300 permutations, the extracted-path veto ratio is not distinguishable from the shuffled controls (maximum |z| 0.220); none modes meet the descriptive point-gap criterion and erdos_renyi, rewire_degree_preserving, sign_permute, sign_permute_weight_matched, weight_permute are indeterminate. The planted-cycle study of §S2.3 does not establish power for this endpoint, so the result cannot rule out a wiring effect. The comparator ratio is undefined at this dose, not zero. -**Selectivity landscape.** The split is by mechanism rather than potency: chlorpyrifos_oxon, ddt, deltamethrin, permethrin are amplified by the circuit and acetamiprid, acetylcholine, clothianidin, imidacloprid, nicotine, nitenpyram buffered, subject to the threshold caveat of §3.4. Compounds never reaching the circuit threshold on either cut (chlordimeform, dieldrin, fipronil, gaba, ivermectin, picrotoxin, spinosad, sulfoxaflor, thiamethoxam) are reported **unscored** rather than buffered, and one of them is unscored for a substrate reason rather than a pharmacological one: chlordimeform's only insect mechanism acts through octopamine, and neither cut contains an octopaminergic cell. +**Selectivity landscape.** The split is by mechanism rather than potency: ddt, deltamethrin are amplified by the circuit and acetamiprid, acetylcholine, clothianidin, imidacloprid, nicotine, nitenpyram buffered, subject to the threshold caveat of §3.4. Compounds never reaching the circuit threshold on either cut (chlordimeform, dieldrin, fipronil, gaba, ivermectin, permethrin, picrotoxin, thiamethoxam) are reported **unscored** rather than buffered, and one of them is unscored for a substrate reason rather than a pharmacological one: chlordimeform's only insect mechanism acts through octopamine, and neither cut contains an octopaminergic cell. -**Literature concordance** is reported per entry with its *n* (T4) and never as a bare mean: of 8 evaluable entry–assay combinations, 4 order exactly two compounds, where Spearman's ρ can only be ±1 whatever the model does, and over the 4 entries with more than two the mean ρ is 0.756. As a concordance count the result is 8 concordant against 0 discordant orderings (two-sided binomial *p* = 0.008); 16 combinations are skipped for want of coverage and 2 inversions are recorded rather than repaired (nitenpyram_rank_inverted, clothianidin_vs_imidacloprid_inverted). Of the 6 distinct published orderings the library can cover, 1 comes from a publication the library already cites for a compound in that ordering and is therefore *not* out-of-sample, while 5 are source-disjoint. None is an independent test of the *circuit* model. The mixture arm is a **limited qualitative concordance check** (3 of 3 verdicts reproduced, reference organism Apis mellifera, 2 partners same-class substitutes, §S4). +**Literature concordance** is reported per entry with its *n* (T4) and never as a bare mean: of 8 evaluable entry–assay combinations, 4 order exactly two compounds, where Spearman's ρ can only be ±1 whatever the model does, and over the 4 entries with more than two the mean ρ is 0.756. As a concordance count the result is 8 concordant against 0 discordant orderings (two-sided binomial *p* = 0.008); 16 combinations are skipped for want of coverage and 2 inversions are recorded rather than repaired (nitenpyram_rank_inverted, clothianidin_vs_imidacloprid_inverted). Of the 6 distinct published orderings the library can cover, 1 comes from a publication the library already cites for a compound in that ordering and is therefore *not* out-of-sample, while 5 show no detected overlap among recorded DOI/PMID identifiers; missing identifiers leave source independence unresolved. None is an independent test of the *circuit* model. The mixture arm is a **limited qualitative concordance check** (3 of 3 verdicts reproduced, reference organism Apis mellifera, 2 partners same-class substitutes, §S4). diff --git a/papers/SUPPLEMENT.md.in b/papers/SUPPLEMENT.md.in index e0255dd..e223ba1 100644 --- a/papers/SUPPLEMENT.md.in +++ b/papers/SUPPLEMENT.md.in @@ -7,24 +7,24 @@ every {{key}} is substituted from `papers/results.json`. --> # Supplementary material -**Supplement to:** *Which Inputs Does a Network Simulation Depend On? Typed Evidence Propagation and Input-Dependence Testing, Evaluated on Cuts of a Synapse-Resolution* Drosophila *Connectome*. +**Supplement to:** *Auditing Input Dependence in Literature-Parameterised Network Simulations: A Connectome Case Study*. -FlyLab {{flylab_version}}, commit `{{git_sha}}`, library SHA-256 `{{library_sha256}}`, map `{{map_id}}`. Reference numbers `[n]` refer to the manuscript's reference list. Every number here is generated by `scripts/reproduce_paper.py` from committed artifacts; none was measured in a living animal. +FlyLab {{flylab_version}}, library SHA-256 `{{library_sha256}}`, map `{{map_id}}`. An internal code revision is recorded in the provenance manifest; a public revision identifier is omitted from blind-review text. Reference numbers `[n]` use the 76-entry source-paper bibliography. The standard script regenerates core model values from the current parameter library; the larger scale-study values are imported from a saved record whose 5k–50k cuts are absent here. No value was measured in a living animal. The library source-value audit remains incomplete. ## S1. Mechanism rules: the gain patch and its rationale -`flylab/pharm/mechanisms.py` is the single source of truth mapping engagement rows onto circuit gains; no assay carries its own copy, which is what makes Table T2 a truthful description of the software. Gains are dimensionless multipliers on the synaptic weight of a transmitter class (or, for `g_nav`, on node excitability); 1.0 is vehicle, and every gain is floored at 0.05 because a network with a gain of exactly zero is not informative. The {{mechanism_rules}} rules are listed in full in T2. The four shipped since v0.4 may not be changed silently: +`flylab/pharm/mechanisms.py` is the single source of truth mapping engagement rows onto circuit gains; no assay carries its own copy. Table T2 therefore describes the implemented rules. Gains are dimensionless multipliers on the synaptic weight of a transmitter class (or, for `g_nav`, on node excitability); 1.0 is vehicle, and every gain is floored at 0.05 because a network with a gain of exactly zero is not informative. The {{mechanism_rules}} rules are listed in full in T2. The four core rules are: -* **Insect nAChR, activating:** `g_ach = max(0.05, 1 + 0.4θ − 1.6θ²)`. Rises at low engagement and falls at high engagement, encoding excitation followed by desensitisation and depolarisation block. At θ = 0.99 it is at the floor, which is why a saturating dose of a nicotinic agonist silences cholinergic transmission in the model rather than merely reducing it — and, as §3.4 of the manuscript shows, why the suppression conclusion is a property of *this* rule. +* **Insect nAChR, activating:** `g_ach = max(0.05, 1 + 0.4θ − 1.6θ²)`. Rises at low engagement and falls at high engagement, encoding excitation followed by desensitisation and depolarisation block. At θ = 0.99 it reaches the model's gain floor of 0.05, sharply reducing cholinergic transmission. As §3.4 of the manuscript shows, the suppression predicate depends on this rule. * **Insect nAChR, blocking:** `g_ach = max(0.05, 1 − θ)`. * **Insect RDL, blocking:** `g_gaba = max(0.05, 1 − θ)`. Blocking the GABA-gated chloride channel removes inhibition. * **Insect RDL, activating:** `g_gaba = max(0.05, 1 + 0.4θ)`. -Four families added in v0.5 share the shape and state a rationale each: GluCl activation uses a larger coefficient (`1 + 0.8θ`) than RDL because avermectin opening is effectively irreversible [38]; AChE inhibition raises an ambient cholinergic tone (`ach_tone = 1 + 2.0θ`) multiplying `g_ach` together with the agonist curve, reproducing the organophosphate excitation-then-block picture; Nav modulation scales node excitability globally (`g_nav = 1 + 1.5θ`), since pyrethroids and DDT act on every cell regardless of transmitter [23], [67]; octopamine agonism uses the smallest coefficient (`1 + 0.5θ`) because it is modulatory. +Four additional families use stated modelling rationales: GluCl activation uses a larger coefficient (`1 + 0.8θ`) than RDL because avermectin opening is effectively irreversible [37]; AChE inhibition raises an ambient cholinergic tone (`ach_tone = 1 + 2.0θ`) multiplying `g_ach` together with the agonist curve, reproducing the organophosphate excitation-then-block picture; Nav modulation scales node excitability globally (`g_nav = 1 + 1.5θ`), since pyrethroids and DDT act on every cell regardless of transmitter [22], [66]; octopamine agonism uses the smallest coefficient (`1 + 0.5θ`) because it is modulatory. -**The level-C generic rule.** The ablation ladder's topology-only level is given one number per compound instead of a mechanism-specific gain vector. Until v0.6.1 that number was `{{abl_generic_floor_rule}}`, applied to every transmitter alike — a rule that can only depress, so every one of its {{abl_compounds}} × {{abl_concs}} entries was at most zero and the level could not express disinhibition at all. It is now `{{abl_generic_rule}}`: the same generic magnitude, with one bit of sign read off the mechanism table. {{abl_generic_floor_note}} Both are reported, the old one as `C_topology_only_floor` in T13 and T13b. +**The level-C generic rule.** The ablation ladder gives its topology-only level one signed multiplier per compound rather than a mechanism-specific vector. The direction-aware rule is `{{abl_generic_rule}}`; its magnitude matches the depression-only comparison `{{abl_generic_floor_rule}}`. {{abl_generic_floor_note}} Both are reported for {{abl_compounds}} compounds at {{abl_concs}} concentrations in T13 and T13b. Their difference assesses sensitivity to the generic rule and is not a biological contrast. -**These coefficients and shapes are modelling choices, not measurements.** That is why the manuscript prespecifies a family of {{stab_family_size}} alternatives (§2.4) and re-derives every conclusion under each, and why the variance analysis nominates this transformation as the assumption worth measuring first. The **transmitter sign table** the rate engine iterates is a second set of asserted coefficients and is published as T21: {{sign_asserted}} of the seven magnitudes ({{sign_asserted_transmitters}}) carry no measurement, are members of no specification family, and are not bounded by the label-permuting null, which permutes assignments and leaves the table itself untouched. +**These coefficients and shapes are modelling choices, not measurements.** The manuscript evaluates {{stab_family_size}} gain-rule specifications (§2.4), but the five asserted transmitter sign magnitudes ({{sign_asserted_transmitters}}; T21) and transmitter-defined gain assignment remain fixed. The label-permutation null changes which nodes receive labels; it does not vary the sign magnitudes. Therefore the family does not bound uncertainty from these assumptions. ### S1.1 The specification family @@ -40,14 +40,11 @@ Five shapes — monotone-linear, saturating, weakly biphasic, the shipped biphas | C6 | {{stab_C6_claim}} | {{stab_C6_retained}} | {{stab_C6_lost}} | {{stab_C6_undecidable}} | | C7 | {{stab_C7_claim}} | {{stab_C7_retained}} | {{stab_C7_lost}} | {{stab_C7_undecidable}} | -**What the v0.6 matrix could not have told you.** Two defects made the topology rows of the previous matrix empty rather than merely noisy, and both are fixed here. +**Specification-test scope.** The topology predicates use {{stab_n_shuffles}} shuffles per specification, allowing empirical probabilities to reach α = {{stab_topology_alpha}}. The specification is passed explicitly into the null engine. The C1–C7 rows of T14 are predicates about model outputs under this fixed implementation and selected specification family; they are not biological validation. -1. *The shuffle count could not reject.* The topology predicates were evaluated at 6 shuffles per specification, where the smallest attainable empirical permutation probability is 1/7 = 0.143. `p ≤ 0.05` was arithmetically unattainable, so the predicate fell back on a Gaussian `|z| < 2` and C3's 25/25 was a non-rejection under a test that could not reject. The run behind this document uses {{stab_n_shuffles}} shuffles per specification (resolution {{stab_shuffle_resolution}}), which supports α = {{stab_topology_alpha}} with room to spare. -2. *The specification never reached the null engine.* `robustness.mechanism_spec` rebinds `compute_gains` on the assay modules, but the permutation path resolves gains through `flylab.circuit.rate.compute_gains`, which it did not rebind. Every specification therefore fed the **default** gains to its nulls and produced identical effects, so a uniform retention rate across the family was guaranteed by construction. The dependence engine now accepts the gain patch explicitly (`gains_by_compound`), and the per-specification topology verdicts in T14 are computed under the specification they are labelled with. +The topology predicates are retained in {{stab_C3_retained}}/{{stab_family_size}} and {{stab_C4_retained}}/{{stab_family_size}} cases after Benjamini-Hochberg adjustment of the {{stab_n_structural_tests}} structural component tests (uncorrected counts {{stab_C3_retained_uncorrected}} and {{stab_C4_retained_uncorrected}}). This adjustment does not give a cell-level or predicate-level FDR guarantee. Among C3 cases, {{stab_C3_equivalent}} pass the descriptive point-gap criterion and {{stab_C3_indeterminate}} remain indeterminate; the C4 split is {{stab_C4_equivalent}} / {{stab_C4_indeterminate}}. These counts are conditional on the fixed sign table, assignment rule and common shuffle stream, not independent replications. -Both topology conclusions survive the correction ({{stab_C3_retained}}/{{stab_family_size}} and {{stab_C4_retained}}/{{stab_family_size}} after Benjamini-Hochberg across the {{stab_n_structural_tests}} structural tests of the run; {{stab_C3_retained_uncorrected}} and {{stab_C4_retained_uncorrected}} uncorrected), but a retention rate for a conclusion shaped as a *failure to reject* is only as strong as the equivalence behind it. Of C3's {{stab_C3_retained}}, **{{stab_C3_equivalent}} reach equivalence within the prespecified margin and {{stab_C3_indeterminate}} are merely indeterminate**; the corresponding split for C4 is {{stab_C4_equivalent}} / {{stab_C4_indeterminate}}. Every specification's topology cell is run on the same shuffle stream, so the structural tests are positively dependent rather than independent; Benjamini-Hochberg controls the FDR under positive dependence, but the fractions describe this shuffle stream and are not {{stab_family_size}} independent replications. - -Specifications reversing C1: {{stab_C1_failing}}. Mean SI gap over the nicotinic set under the shipped rule: {{stab_mean_gap_nicotinic_default}} log₁₀ units; averaged over the {{stab_mean_gap_nicotinic_monotone_n}} reversing specifications for which the index is defined: {{stab_mean_gap_nicotinic_monotone}}. **The v0.6 supplement said the monotone rules "buffer at least as much, never less". That is false.** The per-specification range runs {{stab_mean_gap_nicotinic_monotone_max}} to {{stab_mean_gap_nicotinic_monotone_min}}, {{stab_monotone_buffer_less_n}} of them buffer *less* than the shipped rule ({{stab_monotone_buffer_less}}), and the aggregate is carried by the extreme of the range. C5 survives the specifications that reverse C1, but not for the reason the previous version gave. +Specifications reversing C1: {{stab_C1_failing}}. The mean selectivity-index gap in the nicotinic set under the shipped gain rule is {{stab_mean_gap_nicotinic_default}} log₁₀ units; over {{stab_mean_gap_nicotinic_monotone_n}} reversing specifications with a defined index it is {{stab_mean_gap_nicotinic_monotone}}. The per-specification range is {{stab_mean_gap_nicotinic_monotone_max}} to {{stab_mean_gap_nicotinic_monotone_min}}; {{stab_monotone_buffer_less_n}} buffer less than the shipped rule ({{stab_monotone_buffer_less}}). An aggregate gap should not be interpreted as a uniform response across specifications. ## S2. Null-model definitions, and the conventions around them @@ -59,7 +56,7 @@ Specifications reversing C1: {{stab_C1_failing}}. Mean SI gap over the nicotinic * **`rewire_degree_preserving`** — double-edge swaps, roughly 10 × |E| accepted swaps. Every node keeps its in-degree, out-degree and transmitter; the wiring pattern is destroyed. * **`erdos_renyi`** — the same node and edge counts, endpoints drawn uniformly, weights drawn from the empirical weight list. Only the size of the graph and its census survive. -**Why `sign_permute` is not a transmitter-identity null.** The v0.6 supplement stated that permuting labels preserves the transmitter histogram "and therefore the global excitation/inhibition balance, preserved *exactly*". That is wrong, and the error was load-bearing: out-strength is heterogeneous across nodes, so the *weighted* balance moves. Measured (T23), the cholinergic share of total outgoing synaptic weight on the `named` cut is {{bal_real_ach_share}} on the real graph against {{bal_plain_ach_mean}} ± {{bal_plain_ach_sd}} across {{bal_n}} permutations, range {{bal_plain_ach_min}}–{{bal_plain_ach_max}}, putting the real graph at the {{bal_plain_ach_percentile}}th percentile of its own null. Because a gain patch acts *through* transmitter identity, the mode moves the drug's target set and the sign matrix at once, and the extra variance makes it conservative. The weight-matched null holds every tracked share to within {{bal_matched_max_deviation}}. Substituting it does not change imidacloprid's permutation probability but it shrinks the equivalence gap from {{dep_named_imidacloprid_gap_sign_permute}} Hz to {{dep_named_imidacloprid_gap_sign_permute_weight_matched}} Hz against a margin of {{dep_named_imidacloprid_delta}} Hz. +**Transmitter-label null.** Plain label permutation preserves class counts but changes their weighted outgoing shares because node out-strength varies. In the `named` cut, the cholinergic outgoing-weight share is {{bal_real_ach_share}} on the real graph and {{bal_plain_ach_mean}} ± {{bal_plain_ach_sd}} across {{bal_n}} permutations (range {{bal_plain_ach_min}}–{{bal_plain_ach_max}}), with the observed share at the {{bal_plain_ach_percentile}}th percentile of that null. It changes both the compound target set and the signed matrix. The matched variant constrains each tracked weighted share to within {{bal_matched_max_deviation}}. The imidacloprid median point gap changes from {{dep_named_imidacloprid_gap_sign_permute}} to {{dep_named_imidacloprid_gap_sign_permute_weight_matched}} rate-model units against margin {{dep_named_imidacloprid_delta}}. These nulls still hold the transmitter sign magnitudes fixed. Every mode preserves the node set, the node order and the seed block, and never creates self-loops or duplicate edges. The effect is always `treated − vehicle` with **both arms on the same graph**. The reported probability is the empirical two-sided permutation *p*, `(k+1)/(n+1)`, with resolution `1/(n+1)`; on a landscape the same `n` shuffled graphs are reused by every cell (a paired design), and shuffle *i* depends only on `(seed, i)`, so each cell's numbers are identical to running its profile alone. @@ -67,43 +64,43 @@ Every mode preserves the node set, the node order and the seed block, and never ### S2.1 The three-way verdict and the equivalence margin -Failing to reject a null is not evidence of equivalence, and none of the software's return values says it is. Each mode gets exactly one of `distinguishable`, `equivalent_within_tolerance` or `indeterminate`, against a margin δ prespecified as {{dep_named_imidacloprid_delta_frac}} of the **vehicle readout** of the real graph. The vehicle readout is the right scale because it is in the units of the readout, does not depend on the drug, on the null ensemble or on the number of shuffles, and is the same for every mode of a cell, so an equivalence claim made against it is comparable across the ladder. It is one tenth of the 50 % relative change the selectivity module treats as a circuit response. A cell whose baseline is undefined or zero can form no margin; its non-rejected modes are then `indeterminate`, which is the correct answer rather than a missing one. +Each graph mode receives `distinguishable`, `equivalent_within_tolerance` or `indeterminate` under a margin δ set at {{dep_named_imidacloprid_delta_frac}} of the real graph’s vehicle readout. `equivalent_within_tolerance` requires non-rejection plus an observed point gap below δ; it is a descriptive rule, not a formal equivalence test based on a confidence interval. A zero or undefined baseline yields no usable margin and non-rejected modes are indeterminate. -**The confirmatory profile in full.** For imidacloprid at 1 µM on `named`, at {{dep_n}} permutations against a margin of {{dep_named_imidacloprid_delta}} Hz, the per-mode probabilities and verdicts are: weight permutation *p* = {{dep_named_imidacloprid_p_weight_permute}}, {{dep_named_imidacloprid_verdict_weight_permute}}; degree-preserving rewiring *p* = {{dep_named_imidacloprid_p_rewire_degree_preserving}}, {{dep_named_imidacloprid_verdict_rewire_degree_preserving}}; the weight-matched transmitter null *p* = {{dep_named_imidacloprid_p_sign_permute_weight_matched}}, {{dep_named_imidacloprid_verdict_sign_permute_weight_matched}}; the joint label-permutation null *p* = {{dep_named_imidacloprid_p_sign_permute}}, {{dep_named_imidacloprid_verdict_sign_permute}}; and the Erdős–Rényi control *p* = {{dep_named_imidacloprid_p_erdos_renyi}}, {{dep_named_imidacloprid_verdict_erdos_renyi}}. The resulting necessary level is `{{dep_named_imidacloprid_level}}` with verdict {{dep_named_imidacloprid_level_verdict}}. +**The confirmatory profile in full.** For imidacloprid at 1 µM on `named`, at {{dep_n}} permutations against a margin of {{dep_named_imidacloprid_delta}} rate-model units, the per-mode probabilities and verdicts are: weight permutation *p* = {{dep_named_imidacloprid_p_weight_permute}}, {{dep_named_imidacloprid_verdict_weight_permute}}; degree-preserving rewiring *p* = {{dep_named_imidacloprid_p_rewire_degree_preserving}}, {{dep_named_imidacloprid_verdict_rewire_degree_preserving}}; the weight-matched transmitter null *p* = {{dep_named_imidacloprid_p_sign_permute_weight_matched}}, {{dep_named_imidacloprid_verdict_sign_permute_weight_matched}}; the joint label-permutation null *p* = {{dep_named_imidacloprid_p_sign_permute}}, {{dep_named_imidacloprid_verdict_sign_permute}}; and the Erdős–Rényi control *p* = {{dep_named_imidacloprid_p_erdos_renyi}}, {{dep_named_imidacloprid_verdict_erdos_renyi}}. The resulting necessary level is `{{dep_named_imidacloprid_level}}` with verdict {{dep_named_imidacloprid_level_verdict}}. -The **necessary information level** is the weakest rung the test could not distinguish from the real cut, reported with its verdict. Ranks 2 and 3 are not strictly nested — `weight_permute` keeps transmitter identity and loses the weight-topology pairing, the transmitter nulls the reverse — so a richer model can be distinguishable where a weaker one is not. Such cells are **non-monotone** and are flagged: {{dep_land_non_monotone}} of {{dep_land_cells}} on the `named` cut ({{dep_land_non_monotone_cells}}) and {{dep_land_taste_non_monotone}} of {{dep_land_taste_cells}} on `taste_motor`. Their level must be read as "the cheapest graph model this test cannot tell apart from the real one", never as "everything above it is indistinguishable too". +The software field **necessary information level** reports the lowest-ranked passing degradation and its verdict; its name does not assert biological necessity. Ranks 2 and 3 are not strictly nested — `weight_permute` keeps transmitter identity and loses the weight-topology pairing, the transmitter nulls the reverse — so a richer model can be distinguishable where a weaker one is not. Such cells are **non-monotone** and are flagged: {{dep_land_non_monotone}} of {{dep_land_cells}} on the `named` cut ({{dep_land_non_monotone_cells}}) and {{dep_land_taste_non_monotone}} of {{dep_land_taste_cells}} on `taste_motor`. The reported level does not imply that richer degradations also pass. ### S2.2 Effect floor, multiplicity, and confirmatory versus exploratory -**Effect floor.** An absolute floor of 1e-6 Hz against a ~6.6 Hz baseline admits effects at numerical-noise level, so the effective floor is `max(absolute, 1 % × |vehicle readout|)`. Landscapes report class counts under both. On the `named` cut the relative floor moves {{dep_land_below_relative_floor}} cells into `no-effect` ({{dep_land_below_relative_floor_cells}}), changing the counts from {{dep_land_topology_dependent_absfloor}} / {{dep_land_composition_dominated_absfloor}} / {{dep_land_no_effect_absfloor}} to {{dep_land_topology_dependent}} / {{dep_land_composition_dominated}} / {{dep_land_no_effect}}. +**Effect floor.** The absolute floor of 1e-6 rate-model units is below relevant baseline magnitudes, so the effective floor is `max(absolute, 1 % × |vehicle readout|)`. Landscapes report both. On `named`, the relative floor moves {{dep_land_below_relative_floor}} cells to `no-effect` ({{dep_land_below_relative_floor_cells}}). Counts under the absolute floor alone are {{dep_land_topology_dependent_absfloor}} / {{dep_land_composition_dominated_absfloor}} / {{dep_land_no_effect_absfloor}}; after the relative floor they are {{dep_land_topology_dependent_raw}} / {{dep_land_composition_dominated_raw}} / {{dep_land_no_effect_raw}} before multiplicity adjustment and {{dep_land_topology_dependent}} / {{dep_land_composition_dominated}} / {{dep_land_no_effect}} afterward. -**Multiplicity.** A landscape is a screen. Benjamini-Hochberg is applied across the structural tests of the run ({{dep_land_n_structural_tests}} on `named`, {{dep_land_taste_n_structural_tests}} on `taste_motor`), the class is decided on the adjusted probabilities, and both counts are reported. Because an empirical permutation probability cannot fall below `1/(n+1)`, the result also reports how many tests must sit at that floor before anything can be rejected, so that a corrected count of zero is never mistaken for a finding (`fdr_can_reject` = {{dep_land_fdr_can_reject}} here). +**Multiplicity.** The landscape is exploratory. Benjamini-Hochberg adjusts its declared structural component-test probabilities ({{dep_land_n_structural_tests}} on `named`, {{dep_land_taste_n_structural_tests}} on `taste_motor`), and classes are recomputed from those adjusted values. This does not imply an FDR guarantee for the derived cell labels. The report also states the empirical permutation resolution and whether any adjusted component test can cross α at that resolution (`fdr_can_reject` = {{dep_land_fdr_can_reject}}). -**Confirmatory versus exploratory.** The imidacloprid and fipronil profiles at *n* ≥ 1000 are prespecified planned comparisons and carry no correction; everything else is exploratory and says so in its warnings. Fipronil at 1 µM is the clean illustration: confirmatory *p* = {{dep_named_fipronil_min_structural_p}}, but inside the exploratory landscape its adjusted probability is {{dep_land_fipronil_q}} and the cell does not survive correction ({{dep_land_fipronil_class}} corrected against {{dep_land_fipronil_class_raw}} raw). +**Planned versus exploratory comparisons.** The imidacloprid and fipronil profiles at *n* ≥ 1000 are planned comparisons reported without multiplicity adjustment. Landscape screening is exploratory. For fipronil at 1 µM, the planned component-test *p* is {{dep_named_fipronil_min_structural_p}}; in the landscape the adjusted component-test value is {{dep_land_fipronil_q}} and the derived class is {{dep_land_fipronil_class}} versus {{dep_land_fipronil_class_raw}} before adjustment. ### S2.3 Does the instrument work? Ground-truth recovery and power -Without this, a negative from the ladder is indistinguishable from an under-powered test. A synthetic cut of {{val_recovery_nodes}} nodes and about {{val_recovery_edges}} edges is built with log-normal weights and a transmitter composition close to the real cut's, and a directed cycle of cholinergic nodes passing through the driven seeds is planted in it at a known strength. The cycle is recurrent cholinergic gain, so a patch that collapses `g_ach` removes an amplification that exists only while the cycle is intact: the drug *contrast*, not merely the rate, depends on the wiring. Strength 0 plants nothing and is the negative control. Nothing here touches the connectome or the compound library. +The planted-cycle synthetic test examines whether the procedure recovers one known wiring-dependent contrast. A synthetic cut of {{val_recovery_nodes}} nodes and about {{val_recovery_edges}} edges is built with log-normal weights and a transmitter composition close to the real cut’s. A directed cholinergic cycle through the driven seeds is planted at known strength; strength 0 is the negative control. This test does not estimate power for arbitrary circuit effects or validate the pharmacological parameter library. -{{val_recovery_statement}} Over the power grid ({{val_power_replicates}} independently generated graphs per cell, planted strengths {{val_power_strengths}}, permutation counts {{val_power_ns}}) the detection rate by strength is {{val_power_by_strength}} and by permutation count {{val_power_by_n}}; the empirical false-positive rate on the unplanted control is {{val_power_control_hits}} of {{val_power_control_runs}} ({{val_power_fpr}}). Effect size dominates: raising the permutation count buys far less than raising the planted strength. Full grid: T22, Fig. F16. +{{val_recovery_statement}} In a finite grid of {{val_power_replicates}} independently generated graphs per cell, planted strengths {{val_power_strengths}} and permutation counts {{val_power_ns}}, detection by strength is {{val_power_by_strength}} and by count {{val_power_by_n}}. The unplanted control has {{val_power_control_hits}} detections in {{val_power_control_runs}} runs ({{val_power_fpr}}). These frequencies are specific to this planted mechanism and grid. **Permutation counts in this document.** {{dep_n}} per mode for the `named` profiles (resolution {{dep_p_floor}}), {{dep_n_taste}} for the `taste_motor` path arm, {{dep_n_landscape}} for each of the {{dep_land_cells}} `named` landscape cells and {{dep_n_landscape_taste}} for each of the {{dep_land_taste_cells}} `taste_motor` cells. The sweep (T11) shows the verdict at α = 0.05 settling from *n* ≈ {{dep_verdict_stable_from_n}}–{{dep_verdict_stable_to_n}}, and a mid-range *p* reaching ±0.02 of its final value between *n* = {{dep_p_within_0p02_from_n}} and {{dep_p_within_0p02_worst_n}}. ### S2.4 Where does the verdict settle with the size of the cut? -The verdict reversed between the two committed cuts, and two cuts are not a trend. `flylab.analysis.scale.dependence_vs_scale` runs the *same* dependence profile (`dependence_profile`, unchanged) across a ladder of nested cuts of {{scale_ladder_sizes}} cells grown from the taste-motor seed set by one fixed rule, so a verdict can be compared across size without the construction changing underneath it. The two committed cuts are built by a different recipe, so the clean scale axis is the `scale_*` rungs alone; a sequence that interleaves the committed cuts mixes a size effect with a construction effect, and `verdict_stability(recipe="scale_ladder")` reports each separately. A verdict counts as *settled* only if it is the same on the two largest cuts run with no change above that point — one rung of agreement is n = 1 and is not stability. +The descriptive dependence labels differ between the two committed cuts. A saved scale-study record contains a nested-cut ladder of {{scale_ladder_sizes}} cells grown by a fixed ranked-BFS recipe. The 5k–50k input cuts are absent from this package, so their recorded results cannot be independently rebuilt here. The two committed cuts use another construction rule, and contrasts between them confound size, density, composition and extraction. Cost is the constraint, not correctness. From the per-shuffle and per-edge constants measured on this machine (T26), a profile at *n* = {{scale_frontier_n_paper}} on the {{scale_top_rung}} rung is {{scale_top_profile_hours}} hours on one core and the landscape over the specification family is {{scale_top_landscape_days}} days per specification; a whole-CNS profile at the weight-5 floor is {{scale_whole_cns_profile_hours}} hours and the dense LIF matrix stops fitting in memory well before it. The permutation count must therefore fall with the cut, and `permutation_budget` refuses to go below the resolution floor `1/α − 1`, so every rung records the *n* it was run at. -The ladder's smallest rung, `scale_1k`, is already the decisive contrast with the two cuts of T19: {{scale_1k_nodes}} nodes, {{scale_1k_edges}} edges, mean degree {{scale_1k_mean_degree}}, in-star {{scale_1k_in_star}} — *fewer* nodes than `named` and a densely recurrent cut nothing like it. +The saved ladder's smallest rung, `scale_1k`, provides a descriptive comparison with the committed cuts of T19: {{scale_1k_nodes}} nodes, {{scale_1k_edges}} edges, mean degree {{scale_1k_mean_degree}}, in-star {{scale_1k_in_star}} — fewer nodes than `named`, but a denser cut under a different construction rule. -**Result** (T27; the study record is read from `scale_study.json` beside the paper artifacts, or from `data/derived/` when it is promoted there). Over the {{scale_study_n_cuts}} cuts run ({{scale_study_cuts}}), the composition-dominated verdict survives on {{scale_study_n_composition_dominated}} — {{scale_study_composition_dominated_cuts}} — so {{scale_study_density_not_size}}. Per-rung sequences: {{scale_study_imidacloprid_sequence}}. Fipronil is topology-dependent on every rung. In summary: {{scale_study_statement}} On the clean `scale_*` ladder the verdicts settle: {{scale_study_settled_clean_ladder}}. At {{scale_study_every_null_cuts}} the necessary information level is the real network itself, i.e. the prediction is distinguishable from *every* rung of the ladder. Imidacloprid's relative effect stays within {{scale_study_imidacloprid_relative_range}} of the vehicle readout across the ladder, so the verdict changes while the quantity being measured does not. +**Imported result** (T27, from saved `scale_study.json`). Across {{scale_study_n_cuts}} recorded cuts ({{scale_study_cuts}}), the composition-dominated label occurs on {{scale_study_n_composition_dominated}} ({{scale_study_composition_dominated_cuts}}). Per-rung imidacloprid labels are {{scale_study_imidacloprid_sequence}}; the record labels fipronil topology-dependent on each rung. The record summaries are {{scale_study_statement}} and {{scale_study_density_not_size}}. It reports {{scale_study_every_null_cuts}} cuts where imidacloprid is distinguishable from all three ranked edge/wiring degradations (`weight_permute`, `rewire_degree_preserving`, `erdos_renyi`). The orthogonal transmitter-label checks are separate: plain `sign_permute` is indeterminate at `scale_10k`. The imidacloprid relative effect lies within {{scale_study_imidacloprid_relative_range}} of the vehicle readout, and the clean-ladder summary is {{scale_study_settled_clean_ladder}}. These are conditional observations from the saved run; the absent larger input cuts prevent independent regeneration of all rungs. -**Two caveats, both reported rather than absorbed.** The permutation budget falls with the edge count ({{scale_study_n_range}} draws, resolution {{scale_study_resolution_range}}): {{scale_study_coarse_note}}, so those rungs confirm a verdict already settled lower down rather than establishing it. And `named` and `taste_motor` are hops-limited neighbourhoods while the `scale_*` rungs are nested ranked-BFS cuts, so a difference *between* the two families confounds scale with construction; differences *within* the `scale_*` rungs do not. +**Scale-study limits.** Permutation counts decrease as edge count grows ({{scale_study_n_range}} draws; resolutions {{scale_study_resolution_range}}), and {{scale_study_coarse_note}}. The `named` and `taste_motor` cuts are hops-limited whereas the `scale_*` ladder is ranked-BFS. No comparison here identifies recurrence, density or scale as an isolated cause of the class change. ## S3. Runtimes and their calibration -**Rate runtime.** `flylab/circuit/rate.py` holds a dense signed weight matrix `W[j,i] = sign(nt_i) × synapses(i→j)`, with the signs of Table T21. **That matrix is then row-normalised**: `W_signed = W / max(Σ_i |W[j,i]|, 1)`, one denominator per postsynaptic cell. A gain patch afterwards scales the columns by the presynaptic transmitter's gain, so `W_eff = W_signed · diag(g)` and **each cell's recurrent input is a composition-weighted average of its presynaptic gains, normalised by its own total absolute input**. The update is a clipped leaky iteration `r ← clip((1 − α)·r + α·g_nav·(d + W_eff·r), 0, r_max)` with α = 0.3, 80 steps, `r_max` = 300 Hz and *d* the per-node external drive in Hz. It is deterministic, and one evaluation of the {{named_nodes}}-node graph costs about 50 ms with single-threaded BLAS — which is what makes {{dep_n}}-permutation profiles and an {{dep_land_cells}}-cell landscape affordable at all. +**Rate runtime.** `flylab/circuit/rate.py` constructs `W[j,i] = sign(nt_i) × synapses(i→j)` and row-normalises it as `W_signed = W / max(Σ_i |W[j,i]|, 1)` before applying presynaptic gains: `W_eff = W_signed · diag(g)`. The clipped iteration is `r ← clip((1 − α)·r + α·g_nav·(d + W_eff·r), 0, r_max)` with α = 0.3 and 80 steps. Its numerical readout is in **rate-model units**, not calibrated firing hertz; drive and clipping parameters are set on the model scale. This deterministic runtime takes about 50 ms for one {{named_nodes}}-node evaluation with single-threaded BLAS. The normalisation is an **asserted modelling choice**, not a measurement, and it is consequential for RQ2: a readout in which each cell averages its presynaptic gains builds part of "composition-dominated" into itself. Three modes are implemented and swept in §3.3 (T25): @@ -117,19 +114,19 @@ The composition-versus-full rank correlation at 1 µM is {{abl_rho_composition_r **LIF runtime.** `flylab/circuit/lif.py` is a current-based leaky integrate-and-fire network with the parameters of Shiu *et al.* [7]: τ_m = 20 ms, V_rest = −52 mV, V_th = −45 mV, V_reset = −52 mV, refractory 2.2 ms, exponential synaptic current with τ_s = 5 ms, forward Euler at dt = 0.1 ms. A presynaptic spike on edge *i*→*j* injects `w_scale × synapses × sign(nt_i) × gain(nt_i)` mV into *j*. -Two choices are modelling decisions, not literature values. *The synaptic scale is a one-time calibration scan*: `w_scale` = 0.05 mV per synapse and a drive weight of 0.5 mV per event were chosen once, by scanning, so that vehicle MN9 cells fire in a plausible band under the standard {{veto_sugar_hz}} Hz seed drive. No recording was used and no error was minimised against data, so this does not consume the one fitted parameter the project allows — but it is a tuned constant. *A background Poisson drive is required, and it is a choice*: excising a 1-hop neighbourhood removes every input from outside it, so every non-seed cell receives an independent Poisson drive at 0.65 × the seed rate, standing in for the amputated inputs. This is the most consequential free choice in the LIF arm; it sets the operating point and it is why the two engines report sugar-driven MN9 at {{veto_rate_vehicle_sugar}} Hz (rate) against {{veto_lif_vehicle_sugar}} Hz (LIF). +Two LIF choices are modelling decisions, not measured values. The synaptic scale `w_scale` = 0.05 mV per synapse and drive weight 0.5 mV per event were selected by a one-time scan, without fitting to recordings. Background Poisson drive at 0.65 × seed rate is supplied to non-seed cells to stand in for inputs omitted by the 1-hop cut. It sets the LIF operating point. The rate-model output is {{veto_rate_vehicle_sugar}} rate-model units; the LIF output is {{veto_lif_vehicle_sugar}} Hz under the standard {{veto_sugar_hz}} Hz seed drive. The values are not directly calibrated to one another. ## S4. Supporting capabilities These are implemented, tested and used, but are supporting capabilities rather than contributions of this paper. -**Exposure.** `flylab/pharm/exposure.py` maps a dose and route onto *C(t)* and hence engagement(t) in one compartment. Its constants are class placeholders chosen so that the three routes differ in the direction the literature describes, and are fitted to no *Drosophila* ADME dataset because that dataset does not exist: of the seven quantities such a model needs, `data/literature/fly_pharmacokinetics.yaml` records only metabolite identity as well established [52], [54], [55], [58]. Adult haemolymph volume, adult body mass, cuticular penetration rate and any *Drosophila* elimination half-life are `null`; what is recorded is 25 nL recoverable per adult [57], 174 ± 81 nL in third-instar larvae [56], a water-partitioning measurement [53], an ethanol entry that could not supply a rate [65], and an imidacloprid half-life of 4.5–5 h in *honeybee* [58] — the only verified insect half-life for a library compound. Two structural warnings travel with every profile: haemolymph is the lowest-radioactivity compartment for imidacloprid, so a one-compartment haemolymph model under-estimates CNS exposure [59]; and a substantial metabolic route runs through the gut microbiome, so a single-enzyme clearance term is wrong for this compound [54]. +**Exposure.** `flylab/pharm/exposure.py` maps a dose and route onto *C(t)* and hence engagement(t) in one compartment. Its constants are class placeholders chosen so that the three routes differ in the direction the literature describes; they are not fitted to a *Drosophila* ADME dataset. Of the seven quantities such a model needs, `data/literature/fly_pharmacokinetics.yaml` records scattered metabolism observations from *Drosophila*, heterologous cell culture and honeybees [53], [54], [57], but not a complete *Drosophila* exposure series. Adult haemolymph volume, adult body mass, cuticular penetration rate and any *Drosophila* elimination half-life are `null`; what is recorded is 25 nL recoverable per adult [56], 174 ± 81 nL in third-instar larvae [55], a water-partitioning measurement [52], an ethanol entry that could not supply a rate [64], and an imidacloprid half-life of 4.5–5 h in *honeybee* [57] — the only verified insect half-life for a library compound. Two structural warnings travel with every profile: a honeybee study found lower total radioactivity in haemolymph than in other sampled compartments [58], so a one-compartment haemolymph model may miss tissue distribution; the direction and size of any CNS exposure bias are unknown. A substantial metabolic route through the gut microbiome also makes a single-enzyme clearance term incomplete for this compound [53]. -**Genotype.** `flylab/pharm/genotype.py` applies published target-site resistance alleles as multiplicative potency shifts from `data/literature/resistance_alleles.yaml` [13]–[25], enforcing one rule: **a shift applies per compound, never per receptor.** Two published findings make the naive implementation wrong — *Rdl* A301S shifts GABA potency {{geno_gaba_rdl_A301S_nlug_fold}}-fold but fipronil only {{geno_fipronil_rdl_A301S_nlug_fold}}-fold on the expressed planthopper receptor [15], and *para* M918T shifts deltamethrin {{geno_deltamethrin_para_M918T_superkdr_fold}}-fold while leaving DDT at {{geno_ddt_para_M918T_superkdr_fold}}-fold [22]. An allele/compound pair with no sourced number is left unchanged and listed under `unshifted`, never extrapolated. One row is deliberately ugly: *Rdl* A301G in *D. simulans* carries a {{geno_fipronil_rdl_A302G_dsim_fold}}-fold shift derived from a whole-animal resistance ratio, which drives fipronil engagement to {{geno_fipronil_rdl_A302G_dsim_occ}} and is almost certainly an overstatement of the target-site change [19]. The notebook says so. Panel: T9, F9. +**Genotype.** `flylab/pharm/genotype.py` applies published target-site resistance alleles as multiplicative potency shifts from `data/literature/resistance_alleles.yaml` [12]–[24], enforcing one rule: **a shift applies per compound, never per receptor.** Two published findings make the naive implementation wrong — *Rdl* A301S shifts GABA potency {{geno_gaba_rdl_A301S_nlug_fold}}-fold but fipronil only {{geno_fipronil_rdl_A301S_nlug_fold}}-fold on the expressed planthopper receptor [14], and *para* M918T shifts deltamethrin {{geno_deltamethrin_para_M918T_superkdr_fold}}-fold while leaving DDT at {{geno_ddt_para_M918T_superkdr_fold}}-fold [21]. An allele/compound pair with no sourced number is left unchanged and listed under `unshifted`, never extrapolated. One row is deliberately ugly: *Rdl* A301G in *D. simulans* carries a {{geno_fipronil_rdl_A302G_dsim_fold}}-fold shift derived from a whole-animal resistance ratio, which drives fipronil engagement to {{geno_fipronil_rdl_A302G_dsim_occ}} and is almost certainly an overstatement of the target-site change [18]. The notebook says so. Panel: T9, F9. -**Mixtures.** `flylab/pharm/mixtures.py` implements Bliss independence [45] and Loewe additivity [48], [49] as explicit null models, following the response-surface framing of Greco *et al.* [46], and selects between them per pair: Loewe where the components share a receptor key, Bliss where their targets are disjoint. The three-case comparison against the honeybee dataset [51] is a qualitative concordance check and not validation. Table T5. +**Mixtures.** `flylab/pharm/mixtures.py` implements Bliss independence [44] and Loewe additivity [47], [48] as explicit null models, following the response-surface framing of Greco *et al.* [45], and selects between them per pair: Loewe where the components share a receptor key, Bliss where their targets are disjoint. The three-case comparison against the honeybee dataset [50] is a qualitative concordance check and not validation. Table T5. -**Expression weighting.** `flylab/pharm/expression.py` relaxes the assumption that every cell expresses the receptor, using class-level statements from single-cell atlases [39]–[44] mapped by hand onto MaleCNS superclasses, and interpolates each node's gain between 1.0 and the full scalar gain. It is a **sensitivity layer and never the primary model**: none of the cited atlases uses MaleCNS superclass labels, so every mapping row is tagged `cross_atlas_inference`; coverage is {{expr_coverage_overall}} of cells; and receptor expression in identified adult motor neurons is a confirmed gap, so MN9 carries the `unknown` default. Weighting shifts MN9 by {{expr_mn9_delta_hz}} Hz at 1 µM imidacloprid, which we report as the size of the uncertainty rather than as a correction. +**Expression weighting.** `flylab/pharm/expression.py` maps class-level atlas statements [38]–[43] to MaleCNS superclasses by hand and interpolates gains. This is a sensitivity layer: the atlas labels do not match MaleCNS superclasses, every mapping is tagged `cross_atlas_inference`, {{expr_coverage_overall}} of graph-node × receptor-key pairs across the two committed cuts have a coarse class-level annotation; this is annotation availability, not the fraction of cells measured to express a receptor. Identified MN9 motor-neuron expression remains unknown. The rate-model MN9 output shifts by {{expr_mn9_delta_hz}} rate-model units at 1 µM imidacloprid. This is a model sensitivity, not an experimental correction. **Model dose-response.** `flylab/assays/ensemble.py` builds the model's own concentration-response curve, fits a four-parameter Hill function by Nelder–Mead in log-concentration space and bootstraps it. For imidacloprid on the `named` cut the circuit midpoint is {{ic50_M}} with a bootstrap 95 % CI of {{ic50_lo_M}} to {{ic50_hi_M}} (r² = {{ic50_r2}}). The interval does not bracket the point estimate, and that is reported rather than hidden: the bootstrap resamples replicates within each concentration, the lower plateau of the ladder is short, and the midpoint is therefore poorly identified. This is in any case the **model's** circuit midpoint, not a receptor or an animal IC50. The one-at-a-time tornado at 1 µM finds exactly zero span for {{tornado_zero_span_params}} — the engagement is saturated — which the global analysis (§3.5) confirms and quantifies. @@ -138,16 +135,16 @@ These are implemented, tested and used, but are supporting capabilities rather t The following {{pred_n_hypotheses}} hypotheses are written before any experiment and labelled `software_prediction`; `live_result` is `null` in every row and stays null until a real table is imported by hand. They are **prospective, not pre-registered**: a mutable repository is not a registry, and the wording will change only if a tagged release is externally archived with a timestamp. Full table: T3. * **H1** (receptor engagement). At 1 µM, imidacloprid insect nicotinic engagement ≥ 0.9 while vertebrate α4β2 stays ≤ 0.2. Predicted gap {{pred_H1_effect}}. Checkable against published expressed-receptor work rather than against flies. -* **H2** (PER, `protocols/per.md`). The same dose lowers cholinergic drive in the MN9/DNp01 neighbourhood: predicted change in network mean rate {{pred_H2_effect}} Hz. **Specification-dependence caveat.** The *sign* of this prediction is a modelling choice, not a model output: §3.4 shows the same compound at the same engagement takes the network from {{stab_default_vehicle_hz}} Hz to {{stab_default_treated_hz}} Hz under the shipped biphasic rule and from {{stab_monotone_vehicle_hz}} Hz to {{stab_monotone_treated_hz}} Hz under a monotone rule of the same admissible family, and only {{stab_C1_retained}} of {{stab_family_size}} specifications retain the suppression direction. An experimenter should treat H2 as a signed interval over the family rather than as a point prediction, and a live result in either direction falsifies a specification rather than the integration. The control rate this needs could not be sourced; the nearest substitute is 40 % responders at 1 h food deprivation rising to 68 % at 2 h in 463 wild-type flies [61], which also shows that deprivation time is a larger nuisance variable than most effects FlyLab would try to detect. -* **H3** (climbing, `protocols/climbing.md`). Fipronil at 1 µM disinhibits the neighbourhood: predicted change {{pred_H3_effect}} Hz, opposite in sign to H2. **The same caveat applies transitively**: H3 is defined relative to H2, so its contrast inherits H2's specification dependence even though the fipronil arm itself is retained by {{stab_C2_retained}} of {{stab_family_size}} specifications. Negative geotaxis is the best-served endpoint [60], [64], [66]. -* **H4** (negative control). Diazepam engages vertebrate GABA-A and not insect RDL in this library, so it moves no fly readout: predicted effect {{pred_H4_effect}}. A non-zero live effect would falsify either the library row [17] or the assumption that the compound reaches the CNS unmetabolised. -* **H5** (within-animal control). Bitter input vetoes sugar-driven MN9 on the map-extracted path: predicted change {{pred_H5_effect}} Hz relative to sugar alone, matching the published direction [7], [63]. -* **H6**. Fipronil relieves the bitter veto on the map path: predicted rise in the veto ratio {{pred_H6_effect}}. The dependence analysis shows this readout is not distinguishable from its shuffled controls, so a live confirmation would support the pharmacology and **not** the wiring. -* **H7** (selectivity negative control). Picrotoxin perturbs the fly neighbourhood while its receptor selectivity index is ≈ 0: predicted change {{pred_H7_effect}} Hz. Any insect-selectivity claim the tool makes must fail for picrotoxin, and picrotoxin's lack of insect-over-vertebrate selectivity is independently attested [31]. +* **H2** (PER, `protocols/per.md`). At the stated dose, six paired model replicates give a mean change of {{pred_H2_effect}} rate-model units in the MN9/DNp01 neighbourhood (T3). In a separate deterministic profile, the shipped rule gives {{stab_default_vehicle_hz}} to {{stab_default_treated_hz}} rate-model units, while a monotone member gives {{stab_monotone_vehicle_hz}} to {{stab_monotone_treated_hz}}; {{stab_C1_retained}} of {{stab_family_size}} specifications retain suppression. The replicate mean and deterministic endpoint difference use different calculations. These are model outputs for prospective testing, not validated fly effects. The nearest sourced behavioural control concerns food-deprivation time [60], which differs from the model readout. +* **H3** (climbing, `protocols/climbing.md`). The fipronil arm predicts {{pred_H3_effect}} rate-model units in the neighbourhood. Its comparison with H2 inherits H2’s specification dependence; {{stab_C2_retained}} of {{stab_family_size}} gain-rule specifications retain the fipronil predicate. Negative geotaxis provides a prospective assay [59], [63], [65], not current validation. +* **H4** (negative control). Diazepam engages vertebrate GABA-A and not insect RDL in this library, so it moves no fly readout: predicted effect {{pred_H4_effect}}. A nonzero biological effect would challenge the combined model assumptions; identifying its cause would require further experiments. +* **H5** (within-animal control). The extracted sensory path predicts a change of {{pred_H5_effect}} rate-model units when bitter input accompanies sugar drive. Its direction is qualitatively concordant with the published sensory-veto result [7], [62]; the present cut and parameter choices are not an independent validation of that experiment. +* **H6**. Fipronil relieves the bitter veto on the map path: predicted rise in the veto ratio {{pred_H6_effect}}. The dependence analysis shows this readout is not distinguishable from its shuffled controls. A matching biological result would provide concordance with the predicted direction, but these null tests cannot determine whether the wiring contributes. +* **H7** (selectivity negative control). The model predicts a fly-neighbourhood change of {{pred_H7_effect}} rate-model units for picrotoxin while its selectivity index is near 0. This checks the software’s selectivity interpretation; receptor-level cross-species data [30] do not validate the circuit output. Rows where the model's own output disagrees with the prospective direction: {{pred_direction_mismatches}}. -**On sample sizes.** Model-internal intervals ({{pred_n_rep}} replicates of library jitter, drive jitter and the RNG seed) contain no biological variance, so the standardised effect sizes they imply are absurd. T3 therefore caps *d* at 1.0 before computing *n*, which yields {{pred_max_n}} per group for every row. **This is an illustrative planning minimum, not a power calculation**: it means "what a large effect would need", it assumes independent flies where vial-based protocols violate that, and it must not be used to size a study. +**On sample sizes.** Model-internal intervals ({{pred_n_rep}} replicates of library jitter, drive jitter and the RNG seed) contain no biological variance, so effect sizes derived from them cannot support a biological power calculation. T3 therefore caps *d* at 1.0 before computing *n*, which yields {{pred_max_n}} per group for every row. **This is an illustrative calculation, not a study power analysis**: it assumes a large standardised effect and independent flies, whereas vial-based protocols may violate independence; it must not be used to size a study. ## S6. Evidence typing: the full rules @@ -161,9 +158,9 @@ The permitted transformation is a function of two facts — the **parameter type A proxy computes the same Hill expression as its non-proxy sibling but reports itself as a transfer and carries a provenance warning naming what was transferred — the species, the preparation or the chemical class. Asking for a binding occupancy from an EC50, or for any numeric engagement from a placeholder row, raises `EvidenceTypeError`; the library validator rejects a placeholder that carries a value and a non-placeholder that does not. The library currently reports {{ev_library_problems}} validation problems. -**The soundness invariant.** *No numeric engagement reaching a circuit readout may originate from a row whose (parameter type, evidence distance) pair does not admit one.* `tests/test_pharm_evidence.py` checks it exhaustively rather than at a call site: over every row of the shipped library and every entry point that can produce an engagement, and with a companion check that dropping the inadmissible rows changes no gain vector. Mutating the rule table makes the test fail with a message naming the row and the pair. This, rather than the enum, is what the manuscript claims as a computing contribution. +**The admission invariant.** *No numeric engagement reaching an enumerated circuit-readout path may originate from a row whose (parameter type, evidence distance) pair does not admit one.* `tests/test_pharm_evidence.py` checks every shipped row and tested entry point, verifies that dropping inadmissible rows leaves the tested gain vectors unchanged, and detects a rule-table mutation. This tests software admission logic; it does not verify that every admitted number matches its cited assay. -**Census.** {{ev_rows}} rows over {{ev_compounds}} compounds and {{ev_receptor_keys}} receptor keys (Table T10, Figure F11). Parameter types: {{ev_param_EC50}} EC50, {{ev_param_IC50}} IC50, {{ev_param_Kd}} Kd, {{ev_param_unknown}} unknown. Evidence distances: {{ev_dist_E0}} E0, {{ev_dist_E1}} E1, {{ev_dist_E2}} E2, {{ev_dist_E3}} E3, {{ev_dist_E4}} E4. Engagement models: {{ev_model_binding_occupancy}} `binding_occupancy`, {{ev_model_binding_engagement_proxy}} `binding_engagement_proxy`, {{ev_model_functional_engagement}} `functional_engagement`, {{ev_model_functional_engagement_proxy}} `functional_engagement_proxy`, {{ev_model_not_modelled}} `not_modelled`. Rows entitled to a binding-occupancy model: {{ev_binding_row_names}} [77]. Rows demoted to a binding proxy by distance: {{ev_binding_proxy_row_names}}. +**Census.** {{ev_rows}} rows over {{ev_compounds}} compounds and {{ev_receptor_keys}} receptor keys (Table T10, Figure F11). Parameter types: {{ev_param_EC50}} EC50, {{ev_param_IC50}} IC50, {{ev_param_Kd}} Kd, {{ev_param_unknown}} unknown. Evidence distances: {{ev_dist_E0}} E0, {{ev_dist_E1}} E1, {{ev_dist_E2}} E2, {{ev_dist_E3}} E3, {{ev_dist_E4}} E4. Engagement models: {{ev_model_binding_occupancy}} `binding_occupancy`, {{ev_model_binding_engagement_proxy}} `binding_engagement_proxy`, {{ev_model_functional_engagement}} `functional_engagement`, {{ev_model_functional_engagement_proxy}} `functional_engagement_proxy`, {{ev_model_not_modelled}} `not_modelled`. Rows entitled to a binding-occupancy model: {{ev_binding_row_names}} [76]. Rows demoted to a binding proxy by distance: {{ev_binding_proxy_row_names}}. ## S7. Reproduction details @@ -184,7 +181,7 @@ A proxy computes the same Hill expression as its non-proxy sibling but reports i | dose-response bootstrap | {{ic50_n_boot}} resamples | 40 | | prediction replicates | {{pred_n_rep}} | 2 | -The middle column is read from `results.json`, so it describes the run that rendered this document rather than an aspiration. `--fast` changes statistical effort only, never the model: point estimates that do not depend on replicate or permutation count are identical, and a `--fast` *p* is coarser. Two tests guard the committed record: it must not come from a `--fast` run, and — new in this revision — **every statistical parameter it records must equal the shipped default**, because the v0.6 record was generated at a shuffle count the code no longer used and nothing detected it. +The middle column is read from `results.json`, so it describes the run that rendered this document rather than an aspiration. `--fast` changes statistical effort only, never the model: point estimates that do not depend on replicate or permutation count are identical, and a `--fast` *p* is coarser. Two tests guard the committed record: it must not come from a `--fast` run, and every recorded statistical parameter must match the shipped default. Two of the analyses are started in worker processes when the run begins and joined at their own step, because they dominate the wall clock and depend on nothing the other steps produce. `--jobs` controls that and the permutation workers; results are identical at any value. @@ -200,7 +197,7 @@ These are summarised in §4 and given here at length. **(ii) The row normalisation is load-bearing** (§S3, §3.3). -**(iii) Receptor effects are applied across transmitter-defined synapses.** A gain on `g_ach` scales every cholinergic synapse, whether or not that synapse expresses the receptor the compound binds. This is the assumption the expression layer relaxes, and it is reported as a bound rather than an improvement: coverage is {{expr_coverage_overall}} of cells, every mapping row is a cross-atlas inference, and receptor expression in identified adult motor neurons — MN9's class — is a confirmed gap in the atlas literature, so the paper's principal readout carries the default weight. Weighting shifts MN9 by {{expr_mn9_delta_hz}} Hz at 1 µM imidacloprid. +**(iii) Receptor effects are applied across transmitter-defined synapses.** A gain on `g_ach` scales every cholinergic synapse, whether or not that synapse expresses the receptor the compound binds. This is the assumption the expression layer relaxes, and it is reported as a sensitivity analysis: {{expr_coverage_overall}} of graph-node × receptor-key pairs across the two committed cuts have a coarse class-level annotation; this is not a fraction of cells measured to express a receptor. Every mapping row is a cross-atlas inference. We reprocessed the four adult VNC count matrices in GEO GSE141807 into gene-level and target-family detection fractions after the published QC filters; the summary is VNC-wide and contains no cell-class assignments. Three replicate cell counts match the paper; one retains two extra cells under the published thresholds, an unresolved discrepancy recorded with the analysis. Single-cell dropout means a zero count is not evidence of absence. These data do not identify MN9, a proboscis motor neuron whose arbors are in the brain's subesophageal zone, so the principal readout still carries the default weight. Weighting shifts the rate-model MN9 readout by {{expr_mn9_delta_hz}} rate-model units at 1 µM imidacloprid. **(iv) Hops-limited cuts.** A 1-hop neighbourhood with a weight floor is a fraction of MN9's real input; missing inhibition can flip the sign of any readout, and the disinhibition readout is bounded by the GABA share that survived the cut. Growing the cut is deliberately constrained: past about 1 MB of JSON the weight floor is raised rather than hops added, because a 2-hop cut of a motor neuron in a 25-million-edge CNS is most of the animal. §3.2 is the quantitative demonstration of how much the choice of cut matters — the dependence verdict itself reverses between the two committed cuts. @@ -214,10 +211,10 @@ These are the domain-side results the method is demonstrated on. They are suppor **The evidence census in full.** Of {{ev_rows}} rows over {{ev_compounds}} compounds and {{ev_receptor_keys}} receptor keys, {{ev_param_EC50}} carry an EC50, {{ev_param_IC50}} an IC50 and {{ev_param_Kd}} a binding constant. By evidence distance the split is {{ev_dist_E0}} / {{ev_dist_E1}} / {{ev_dist_E2}} / {{ev_dist_E3}} / {{ev_dist_E4}} rows at E0–E4. At the 1 µM working concentration the two headline compounds engage their insect targets at {{imi_insect_occ}} and {{fip_insect_occ}} against {{imi_vert_occ}} and {{fip_vert_occ}} at the vertebrate counterparts ({{imi_ec50_ratio}}-fold and {{fip_ec50_ratio}}-fold; {{window_log10_imidacloprid}} log₁₀ units of window for the first). Placeholder rows contribute nothing: the negative-control row reports "{{dia_insect_rdl_occ}}". -**Sensory-veto readout.** Driving the {{veto_n_sweet_grn}} sweet GRN body IDs at {{veto_sugar_hz}} Hz and reading MN9 reproduces the published direction of a bitter veto [7], [63] on both runtimes ({{veto_rate_vehicle_ratio}} rate, {{veto_lif_vehicle_ratio}} LIF in vehicle; {{veto_rate_fipronil_ratio}} and {{veto_lif_fipronil_ratio}} under fipronil). The two runtimes agree on **direction only** — sugar-driven MN9 is {{veto_rate_vehicle_sugar}} Hz on the rate engine against {{veto_lif_vehicle_sugar}} Hz on the LIF engine, a difference following from the LIF background drive (§S3) — and we do not present that agreement as cross-model validation. +**Sensory-veto readout.** Driving {{veto_n_sweet_grn}} sweet GRN IDs at {{veto_sugar_hz}} Hz produces the published *direction* of bitter veto [7], [62] in both runtimes (vehicle ratios {{veto_rate_vehicle_ratio}} rate model and {{veto_lif_vehicle_ratio}} LIF; fipronil ratios {{veto_rate_fipronil_ratio}} and {{veto_lif_fipronil_ratio}}). This is qualitative concordance only. Sugar-driven MN9 is {{veto_rate_vehicle_sugar}} rate-model units versus {{veto_lif_vehicle_sugar}} Hz in the LIF network; these numbers have different calibration status and are not cross-model validation. -**A properly powered negative.** At {{dep_n_taste}} permutations the veto ratio on the extracted sensory path is not distinguishable from any shuffled control (max |z| = {{dep_taste_fipronil_max_abs_z}}), equivalent within tolerance for {{dep_taste_fipronil_equivalent_modes}} and indeterminate for {{dep_taste_fipronil_indeterminate_modes}}. Because §3.1 characterises the instrument's power, this is reportable as a negative rather than as an absence of evidence. The comparator compound has no veto ratio at this dose at all — the readout is *undefined*, not zero, and is reported as undefined. +**Path-null result.** At {{dep_n_taste}} permutations, the extracted-path veto ratio is not distinguishable from the shuffled controls (maximum |z| {{dep_taste_fipronil_max_abs_z}}); {{dep_taste_fipronil_equivalent_modes}} modes meet the descriptive point-gap criterion and {{dep_taste_fipronil_indeterminate_modes}} are indeterminate. The planted-cycle study of §S2.3 does not establish power for this endpoint, so the result cannot rule out a wiring effect. The comparator ratio is undefined at this dose, not zero. **Selectivity landscape.** The split is by mechanism rather than potency: {{landscape_amplify_compounds}} are amplified by the circuit and {{landscape_buffer_compounds}} buffered, subject to the threshold caveat of §3.4. Compounds never reaching the circuit threshold on either cut ({{landscape_unscored_compounds}}) are reported **unscored** rather than buffered, and one of them is unscored for a substrate reason rather than a pharmacological one: chlordimeform's only insect mechanism acts through octopamine, and neither cut contains an octopaminergic cell. -**Literature concordance** is reported per entry with its *n* (T4) and never as a bare mean: of {{rank_evaluated}} evaluable entry–assay combinations, {{rank_n_degenerate}} order exactly two compounds, where Spearman's ρ can only be ±1 whatever the model does, and over the {{rank_n_informative}} entries with more than two the mean ρ is {{rank_mean_rho_informative}}. As a concordance count the result is {{rank_concordant}} concordant against {{rank_discordant}} discordant orderings (two-sided binomial *p* = {{rank_binomial_p}}); {{rank_skipped}} combinations are skipped for want of coverage and {{rank_known_discrepancies}} inversions are recorded rather than repaired ({{rank_discrepancy_ids}}). Of the {{rank_entries_evaluable}} distinct published orderings the library can cover, {{rank_shared_source}} {{rank_shared_source_verb}} from a publication the library already cites for a compound in that ordering and {{rank_shared_source_be}} therefore *not* out-of-sample, while {{rank_source_disjoint}} {{rank_source_disjoint_be}} source-disjoint. None is an independent test of the *circuit* model. The mixture arm is a **limited qualitative concordance check** ({{mix_n_matching}} of {{mix_n_targets}} verdicts reproduced, reference organism {{mix_species}}, {{mix_n_substituted}} partners same-class substitutes, §S4). +**Literature concordance** is reported per entry with its *n* (T4) and never as a bare mean: of {{rank_evaluated}} evaluable entry–assay combinations, {{rank_n_degenerate}} order exactly two compounds, where Spearman's ρ can only be ±1 whatever the model does, and over the {{rank_n_informative}} entries with more than two the mean ρ is {{rank_mean_rho_informative}}. As a concordance count the result is {{rank_concordant}} concordant against {{rank_discordant}} discordant orderings (two-sided binomial *p* = {{rank_binomial_p}}); {{rank_skipped}} combinations are skipped for want of coverage and {{rank_known_discrepancies}} inversions are recorded rather than repaired ({{rank_discrepancy_ids}}). Of the {{rank_entries_evaluable}} distinct published orderings the library can cover, {{rank_shared_source}} {{rank_shared_source_verb}} from a publication the library already cites for a compound in that ordering and {{rank_shared_source_be}} therefore *not* out-of-sample, while {{rank_source_disjoint}} show no detected overlap among recorded DOI/PMID identifiers; missing identifiers leave source independence unresolved. None is an independent test of the *circuit* model. The mixture arm is a **limited qualitative concordance check** ({{mix_n_matching}} of {{mix_n_targets}} verdicts reproduced, reference organism {{mix_species}}, {{mix_n_substituted}} partners same-class substitutes, §S4). diff --git a/papers/figures/F10_rank_validation.png b/papers/figures/F10_rank_validation.png index 9fcbb35..a5265fb 100644 Binary files a/papers/figures/F10_rank_validation.png and b/papers/figures/F10_rank_validation.png differ diff --git a/papers/figures/F11_evidence_types.png b/papers/figures/F11_evidence_types.png index d61c9ab..42f0935 100644 Binary files a/papers/figures/F11_evidence_types.png and b/papers/figures/F11_evidence_types.png differ diff --git a/papers/figures/F12_dependence_landscape.png b/papers/figures/F12_dependence_landscape.png index 9173a50..e2edb8a 100644 Binary files a/papers/figures/F12_dependence_landscape.png and b/papers/figures/F12_dependence_landscape.png differ diff --git a/papers/figures/F13_ablation.png b/papers/figures/F13_ablation.png index 7c73632..462fe78 100644 Binary files a/papers/figures/F13_ablation.png and b/papers/figures/F13_ablation.png differ diff --git a/papers/figures/F14_conclusion_stability.png b/papers/figures/F14_conclusion_stability.png index 1bf978c..6dc8a79 100644 Binary files a/papers/figures/F14_conclusion_stability.png and b/papers/figures/F14_conclusion_stability.png differ diff --git a/papers/figures/F15_uncertainty_voi.png b/papers/figures/F15_uncertainty_voi.png index c1a7b7e..54050e0 100644 Binary files a/papers/figures/F15_uncertainty_voi.png and b/papers/figures/F15_uncertainty_voi.png differ diff --git a/papers/figures/F16_instrument_validation.png b/papers/figures/F16_instrument_validation.png index e1f8c68..90c9b07 100644 Binary files a/papers/figures/F16_instrument_validation.png and b/papers/figures/F16_instrument_validation.png differ diff --git a/papers/figures/F1_architecture.png b/papers/figures/F1_architecture.png index 6f5386d..a0becd0 100644 Binary files a/papers/figures/F1_architecture.png and b/papers/figures/F1_architecture.png differ diff --git a/papers/figures/F1_architecture.svg b/papers/figures/F1_architecture.svg index 5abcf7e..d0c847c 100644 --- a/papers/figures/F1_architecture.svg +++ b/papers/figures/F1_architecture.svg @@ -6,11 +6,11 @@ - 2026-09-20T13:54:55.572622 + 2026-09-23T17:52:45.617011 image/svg+xml - Matplotlib v3.11.2, https://matplotlib.org/ + Matplotlib v3.10.8, https://matplotlib.org/ @@ -40,7 +40,7 @@ Q 165.564692 11.358 165.564692 16.11 L 165.564692 42.84 Q 165.564692 47.592 173.634231 47.592 z -" clip-path="url(#p20679523bf)" style="fill: #ffffff; stroke: #52514e; stroke-width: 1.1; stroke-linejoin: miter"/> +" clip-path="url(#paaf31f6895)" style="fill: #ffffff; stroke: #52514e; stroke-width: 1.1; stroke-linejoin: miter"/> +" clip-path="url(#paaf31f6895)" style="fill: #cde2fb; stroke: #52514e; stroke-width: 1.1; stroke-linejoin: miter"/> +" clip-path="url(#paaf31f6895)" style="fill: #dbe9fb; stroke: #2a78d6; stroke-width: 1.1; stroke-linejoin: miter"/> +" clip-path="url(#paaf31f6895)" style="fill: #fbe2d6; stroke: #eb6834; stroke-width: 1.1; stroke-linejoin: miter"/> +" clip-path="url(#paaf31f6895)" style="fill: #e6f6ef; stroke: #1baf7a; stroke-width: 1.1; stroke-linejoin: miter"/> +" clip-path="url(#paaf31f6895)" style="fill: #ffffff; stroke: #52514e; stroke-width: 1.1; stroke-linejoin: miter"/> +" clip-path="url(#paaf31f6895)" style="fill: #ffffff; stroke: #52514e; stroke-width: 1.1; stroke-linejoin: miter"/> +" clip-path="url(#paaf31f6895)" style="fill: #cde2fb; stroke: #52514e; stroke-width: 1.1; stroke-linejoin: miter"/> +" clip-path="url(#paaf31f6895)" style="fill: none; stroke: #52514e; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: #52514e; stroke: #52514e; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: none; stroke: #2a78d6; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: #2a78d6; stroke: #2a78d6; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: none; stroke: #eb6834; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: #eb6834; stroke: #eb6834; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: none; stroke: #2a78d6; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: #2a78d6; stroke: #2a78d6; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: none; stroke: #1baf7a; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: #1baf7a; stroke: #1baf7a; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: none; stroke: #52514e; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: #52514e; stroke: #52514e; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: none; stroke: #eb6834; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: #eb6834; stroke: #eb6834; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: none; stroke: #52514e; stroke-width: 1.2; stroke-linecap: round"/> +" clip-path="url(#paaf31f6895)" style="fill: #52514e; stroke: #52514e; stroke-width: 1.2; stroke-linecap: round"/> - - - - - - - - - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + - + - - - - - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - - - - - - - - - - - + + + + + + + + + + + - - - - - - - - + + + + + + + + - - - - - - - - - + + + + + + + + + - - - - - - - - - - + + + + + + + + + + - + - - - - - - - - + - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - + diff --git a/papers/figures/F2_dual_scorecard.png b/papers/figures/F2_dual_scorecard.png index 2d682a3..28ff9e6 100644 Binary files a/papers/figures/F2_dual_scorecard.png and b/papers/figures/F2_dual_scorecard.png differ diff --git a/papers/figures/F3_occupancy_curves.png b/papers/figures/F3_occupancy_curves.png index d107d27..b026119 100644 Binary files a/papers/figures/F3_occupancy_curves.png and b/papers/figures/F3_occupancy_curves.png differ diff --git a/papers/figures/F4_taste_motor_graph.png b/papers/figures/F4_taste_motor_graph.png index 018ad14..5279c98 100644 Binary files a/papers/figures/F4_taste_motor_graph.png and b/papers/figures/F4_taste_motor_graph.png differ diff --git a/papers/figures/F5_bitter_veto.png b/papers/figures/F5_bitter_veto.png index d9626fc..b3ff47f 100644 Binary files a/papers/figures/F5_bitter_veto.png and b/papers/figures/F5_bitter_veto.png differ diff --git a/papers/figures/F6_dependence_profile.png b/papers/figures/F6_dependence_profile.png index 4118b65..ed05e8d 100644 Binary files a/papers/figures/F6_dependence_profile.png and b/papers/figures/F6_dependence_profile.png differ diff --git a/papers/figures/F7_selectivity_landscape.png b/papers/figures/F7_selectivity_landscape.png index ed91073..77e4bff 100644 Binary files a/papers/figures/F7_selectivity_landscape.png and b/papers/figures/F7_selectivity_landscape.png differ diff --git a/papers/figures/F8_dose_response_sensitivity.png b/papers/figures/F8_dose_response_sensitivity.png index b36e8ab..b26526e 100644 Binary files a/papers/figures/F8_dose_response_sensitivity.png and b/papers/figures/F8_dose_response_sensitivity.png differ diff --git a/papers/figures/F9_genotype_panel.png b/papers/figures/F9_genotype_panel.png index b83b413..68e3577 100644 Binary files a/papers/figures/F9_genotype_panel.png and b/papers/figures/F9_genotype_panel.png differ diff --git a/papers/figures/captions.md b/papers/figures/captions.md index 3eca8c6..f2208db 100644 --- a/papers/figures/captions.md +++ b/papers/figures/captions.md @@ -4,32 +4,32 @@ Generated by `scripts/reproduce_paper.py`. Do not edit by hand. **F10_rank_validation.** Left: the model's compound ordering against each published ordering that the teaching library can cover, with Spearman's rho per entry and the number of compounds it orders. 8 entry-assay combinations are evaluable and 16 are skipped because the library cannot cover them; 4 of the evaluable ones order exactly two compounds, where rho can only be +-1, so no bare mean is quoted: over the 4 entries with more than two compounds the mean is 0.756, and as a concordance count the result is 8 concordant against 0 discordant orderings (two-sided binomial p = 0.008). Right: the two inversions the validation module records rather than fixes. Nitenpyram sits near the top of the library's insect nAChR potency ordering and is the least potent neonicotinoid in *Drosophila* whole-animal bioassays, and clothianidin/imidacloprid are ordered the opposite way from the bioassays. Both are receptor-EC50 versus whole-animal-potency differences: the whole-animal number contains uptake, penetration and metabolism, which FlyLab does not model. -**F11_evidence_types.** What the compound library rests on, row by row. Left: the parameter each cited source actually measured. Middle: the ordered evidence distance between that source and this compound at this receptor in this species. Right: the transformation the type system then permits, which is a function of the two. Green marks the 10 on-target (E0) rows and the 1 of them that may use a physical binding-occupancy model; amber marks the proxy models, where the same Hill expression is computed but reported as a transfer with a warning naming the gap; red marks rows that carry no modellable evidence and return N/A rather than a small number. The type system, not a convention, enforces this: asking for an engagement from an untyped row raises rather than returning a value, and a soundness invariant is checked exhaustively over every row and every entry point. +**F11_evidence_types.** What the compound library rests on, row by row. Left: the parameter each library row declares. Some source-to-value links remain under assay-level review. Middle: the declared evidence distance between that source and this compound at this receptor in this species. Right: the transformation the type system then permits, which is a function of the two. Green marks the 7 on-target (E0) rows and the 1 of them that may use a physical binding-occupancy model; amber marks the proxy models, where the same Hill expression is computed but reported as a transfer with a warning naming the gap; red marks rows that carry no modellable evidence and return N/A rather than a small number. The type system, not a convention, enforces this: asking for an engagement from an untyped row raises rather than returning a value, and the admission invariant is checked over shipped rows and enumerated entry points. This does not verify the biological validity of admitted values. -**F12_dependence_landscape.** Every compound of the library at four concentrations, on both committed cuts, classified by the weakest graph model the permutation test could **not** distinguish from the real cut -- never by a model that reproduces the effect. Classes are decided on Benjamini-Hochberg adjusted probabilities across each landscape's structural tests, with a prespecified relative effect floor of 1 % of the vehicle readout. Left, the 1-hop `named` cut: 11 of 84 cells need the wiring pattern, 42 are not distinguishable from any structure-preserving degradation, and 31 do not move the readout at all. Right, the denser `taste_motor` cut: 54 topology-dependent, 0 composition-dominated, 30 no-effect. The composition-dominated majority does not survive the change of substrate, and the `named` cut is an in-star whose degree-preserving rewire is close to the identity (T19). +**F12_dependence_landscape.** Every compound of the library at four concentrations, on both committed cuts, classified by the weakest graph model the permutation test could **not** distinguish from the real cut -- never by a model that reproduces the effect. Classes are decided on Benjamini-Hochberg adjusted probabilities across each landscape's structural tests, with a prespecified relative effect floor of 1 % of the vehicle readout. Left, the 1-hop `named` cut: 0 of 84 cells receive the topology-dependent label, 44 receive composition-dominated and 40 fall below the effect floor. Right, the denser `taste_motor` cut: 44 topology-dependent, 0 composition-dominated, 40 no-effect. The composition-dominated majority does not survive the change of substrate, and the `named` cut is an in-star whose degree-preserving rewire is close to the identity (T19). These are descriptive classes formed from BH-adjusted component tests, without a separate cell-level FDR guarantee. -**F13_ablation.** Each level predicts the same 21 compounds with one layer of information removed, and is scored by how well it reproduces the full model's *ordering* of the library (levels have different units, so values are not comparable; the correlation is signed, so a perfect inversion scores -1 rather than 1). Crosses mark the historical depression-only level-C rule, which applied a multiplier of at most 1.0 to every transmitter alike and so could not express disinhibition: its poor correlation was structural, not a finding about connectomes, and the conclusion drawn from it is withdrawn. With a direction-aware rule level C rises to 0.961 at 1 uM from 0.242. The composition level's agreement with the full model must be read against the matched reference distribution of T24, not on its own. +**F13_ablation.** Each level predicts the same 21 compounds with one layer of information removed, and is scored by rank agreement with the full model's *ordering* of the library (levels have different units, so values are not comparable; the correlation is signed, so a perfect inversion scores -1 rather than 1). Crosses mark the historical depression-only level-C rule, which applied a multiplier of at most 1.0 to every transmitter alike and so could not express disinhibition: its poor correlation follows partly from that restriction. With a direction-aware rule level C rises to 0.968 at 1 uM from 0.161. The short green interval at B shows the 5th–95th percentile of the matched one-gain-per-pseudo-compound B-versus-D reference distribution (T24), with its median dashed. The composition level's agreement with the full model must be read against that distribution. -**F14_conclusion_stability.** Each row is a conclusion this paper could state; each column is one prespecified, admissible way of turning receptor engagement into synaptic gain. Green retains the conclusion, red reverses it, amber means the readout does not exist under that specification. The topology rows are decided on 100 permutations per specification and on Benjamini-Hochberg adjusted probabilities across the 100 structural tests of the run; until v0.6.1 the specification never reached the engine those rows were computed on, so their uniformity was guaranteed by construction rather than measured. Retained by every specification: C2_rdl_disinhibition, C3_imidacloprid_topology_not_distinguishable, C4_fipronil_topology_exceeds, C7_map_bitter_veto. The nicotinic suppression conclusion is reversed by every monotone rule, which is how a result that depends on a modelling choice looks when it is tested rather than asserted; and for a conclusion shaped as a failure to reject, the fraction retained is only as strong as the equivalence behind it (T14's `n_equivalent_within_tolerance` against `n_indeterminate`). +**F14_conclusion_stability.** Each row is a model-output predicate; each column is one tested way of turning receptor engagement into synaptic gain, with transmitter signs and synapse assignment fixed. Green retains the predicate, red reverses it, and amber means the readout does not exist under that specification. The topology rows are decided on 100 permutations per specification and on Benjamini-Hochberg adjusted probabilities across the 100 structural component tests of the run; this is not a cell-level FDR guarantee. Retained by every tested specification: C2_rdl_disinhibition, C3_imidacloprid_topology_not_distinguishable, C4_fipronil_topology_exceeds, C7_map_bitter_veto. The nicotinic suppression predicate reverses under every monotone rule in this family. For a predicate based on non-rejection, T14 separates descriptive point-gap passes from indeterminate cases; neither is a formal equivalence result. -**F15_uncertainty_voi.** Left: the share of the variance of the neighbourhood mean rate attributable to each assumption, first-order and total-order, from a Saltelli cross-sample with Jansen estimators, with bootstrap first-order confidence intervals. A factor is **resolved** only when its interval excludes zero; the starred rows are not, and a factor's index exceeding the noise floor does not resolve it. The dashed line is that floor, defined as the largest magnitude among the *negative* first-order estimates: a true first-order index cannot be negative, so that magnitude is a measured lower bound on the estimator's error, and on this run it is several times the declared null factor's own draw (0.046 set by gain_coef, against lif_seed at -0.007). Right: the same shares scaled back into the readout's variance and mapped onto the experiment that would resolve each assumption; a negative estimate is clipped to zero for the decision value, because a value of information cannot be negative. Green marks a factor that needs no experiment at all, only a re-analysis of data already held. +**F15_uncertainty_voi.** Left: the share of the variance of the neighbourhood mean rate attributable to each assumption, first-order and total-order, from a Saltelli cross-sample with Jansen estimators, with bootstrap first-order confidence intervals. A factor is **resolved** only when its interval excludes zero; the starred rows are not, and a factor's index exceeding the noise floor does not resolve it. The dashed line is that floor, defined as the largest magnitude among the *negative* first-order estimates: a true first-order index cannot be negative, so that magnitude is a measured lower bound on the estimator's error, and on this run it is several times the declared null factor's own draw (0.046 set by gain_coef, against lif_seed at -0.007). Right: the same shares scaled back into the readout's variance and mapped onto the experiment that would resolve each assumption; a negative estimate is clipped to zero for the heuristic score. This is not formal expected value of information. Green marks a factor that needs no experiment at all, only a re-analysis of data already held. -**F16_instrument_validation.** The dependence ladder pointed at a graph whose answer is known in advance. A recurrent cholinergic cycle of known strength is planted in a synthetic cut of the same size, density and transmitter composition as the `named` cut; a gain patch that collapses `g_ach` then removes an amplification that exists only while the cycle is intact, so the drug contrast depends on the wiring by construction. Left: the share of independently generated graphs the ladder classifies topology-dependent, against planted strength and permutation count. Right: the recovery experiment itself, with red marking the strengths the ladder recovered. The unplanted control is the empirical false-positive rate (0 of 18). Without this panel, a negative from the ladder would be indistinguishable from an under-powered test. +**F16_instrument_validation.** The dependence ladder pointed at a graph whose answer is known in advance. A recurrent cholinergic cycle of known strength is planted in a synthetic 250-node cut with about 1500 background edges; it is not matched to the real extract's size or connectivity. A gain patch that collapses `g_ach` then removes an amplification that exists only while the cycle is intact, so the drug contrast depends on the wiring by construction. Left: the share of independently generated graphs the ladder classifies topology-dependent, against planted strength and permutation count. Right: the recovery experiment itself, with red marking the strengths the ladder recovered. The unplanted control yielded a detection count (0 of 18). This is a limited positive-control check, not a general power estimate for the paper's biological endpoints. **F1_architecture.** FlyLab dataflow. A compound and a free concentration enter the occupancy engine, which reads the sourced library (optionally shifted by a genotype, combined with a mixture partner, or driven by an exposure profile). Occupancy splits into two scorecards. The insect side passes through the mechanism table to a gain patch on transmitter classes of a named MaleCNS cut; the vertebrate side is scored at the same dose and never touches a circuit. The circuit runtime (rate or LIF) produces named-cell readouts, which the analysis and validation layers turn into empirical permutation probabilities with their three-way verdicts, selectivity indices, rank correlations and prospective predictions. Everything lands in one notebook JSON with a provenance block. -**F2_dual_scorecard.** Dual scorecard for imidacloprid and fipronil at 1 µM. Bars are Hill occupancy of the insect target (blue) and its vertebrate counterpart (orange) for each of the five receptor pairs declared in `library.yaml:selectivity_pairs`. Hatched bars rest on a class placeholder: the compound has no sourced EC50 at that pair and the row is inert. Imidacloprid reaches 0.991 at insect nAChR against 0.091 at vertebrate a4b2; fipronil reaches 0.985 at insect RDL against 0.476 at native heteromeric vertebrate GABA-A. Teaching-tier EC50s: these are literature-order values, not measured constants. +**F2_dual_scorecard.** Dual scorecard for imidacloprid and fipronil at 1 µM. Bars are Hill-form engagement of the insect target (blue) and its vertebrate counterpart (orange) for each of the five receptor pairs declared in `library.yaml:selectivity_pairs`. Hatched bars rest on a class placeholder: the compound has no sourced EC50 at that pair and the row is inert. Imidacloprid reaches 0.991 at insect nAChR against 0.091 at vertebrate a4b2; fipronil reaches 0.985 at insect RDL against 0.476 at a multi-species vertebrate GABA-related binding-site proxy. These are heterogeneous assay inputs and transfers, not measured fractional occupancy of matched receptors in living animals. -**F3_occupancy_curves.** Hill occupancy against free concentration for the four-compound panel, one insect target (solid, blue) and its vertebrate counterpart (dashed, orange) per compound. The shaded band is the selectivity window between the two EC50s, annotated in log10 units; the dotted vertical line is the 1 µM working concentration used elsewhere in the paper. Ivermectin's window is the GluCl/GlyR pair, deltamethrin's the Nav pair. Curves are the teaching library, not measured concentration-response data. +**F3_occupancy_curves.** Hill-form model engagement against free concentration for the four-compound panel, one insect target (solid, blue) and its vertebrate counterpart (dashed, orange) per compound. The shaded band is the selectivity window between the two EC50s, annotated in log10 units; the dotted vertical line is the 1 µM working concentration used elsewhere in the paper. Ivermectin's window is the GluCl/GlyR pair, deltamethrin's the Nav pair. Curves are the teaching library, not measured concentration-response data. **F4_taste_motor_graph.** Left: the committed MaleCNS v1.0 taste-motor cut (1841 cells, 19066 edges; 1 hop around MN9, DNp01 and the labellar GRN types, plus induced partner-partner edges of at least 10 synapses). Node and edge colour is the predicted consensus transmitter; layout is the deterministic superclass-column layout from `flylab.analysis.layout`. Seeds are marked: sweet GRNs (LB3b/c, up triangles), bitter GRNs (LB1a-d, down triangles), DNp01 (squares) and MN9 (stars). Right: the five strongest sweet and five strongest bitter paths of at most 3 synapses from a labellar GRN to MN9, ranked by the absolute product of signed synapse weights; there are 2595 such sweet and 2372 such bitter paths in total. The sweet/bitter assignment of LB types is a working hypothesis, not a MaleCNS annotation. -**F5_bitter_veto.** MN9 firing on the map-extracted labellar-GRN path, driven through the real MaleCNS sweet (34) and bitter (38) GRN body IDs at 150 Hz, never by driving MN9 directly. Left: leaky rate engine. Right: Shiu-style LIF engine with a background Poisson drive. In vehicle, adding bitter drive takes MN9 to zero on both engines (veto ratio 0.00). Fipronil at 1 µM blocks RDL, raises the sugar-only rate and partially relieves the veto (ratio 0.41 rate, 0.16 LIF). Imidacloprid silences MN9 altogether, so its ratio is undefined rather than zero. Absolute rates differ between engines by construction and are not calibrated to any recording. +**F5_bitter_veto.** MN9 firing on the map-extracted labellar-GRN path, driven through the real MaleCNS sweet (34) and bitter (38) GRN body IDs at 150 Hz, never by driving MN9 directly. Left: leaky rate engine. Right: Shiu-style LIF engine with a background Poisson drive. In vehicle, adding bitter drive takes MN9 to zero on both engines (veto ratio 0.00). Fipronil at 1 µM blocks RDL, raises the sugar-only rate and partially relieves the veto (ratio 0.41 rate, 0.16 LIF). Imidacloprid silences MN9 altogether, so its ratio is undefined rather than zero. Outputs differ between engines by construction and are not calibrated to any recording. -**F6_dependence_profile.** Histograms are the drug effect (treated minus vehicle on the network mean rate) measured on 1000 degraded copies of the 1-hop MN9/DNp01 `named` cut; the red line is the same contrast on the real cut. Columns run in increasing order of destruction: transmitter labels permuted, synapse weights permuted, degree-preserving double-edge rewiring, Erdos-Renyi. Titles carry the empirical two-sided permutation p and the three-way verdict against the prespecified equivalence margin (delta = 0.332 Hz, 5 % of the vehicle readout): `distinguishable` means the test rejected, `equivalent within tolerance` means the gap from the null median is below the margin, and `indeterminate` means neither -- a failure to reject is not evidence that the degraded graph gives the same effect. Plain label permutation is shown because it is informative, but it moves the weighted excitation/inhibition balance as well as transmitter identity (T23), so it is a joint null and not a rung of the ladder. Imidacloprid is composition-dominated (necessary level: degree_sequence, equivalent_within_tolerance); fipronil is topology-dependent (necessary level: wiring_with_weighted_transmitter_balance). +**F6_dependence_profile.** Histograms are the drug effect (treated minus vehicle on the network mean rate) measured on 1000 degraded copies of the 1-hop MN9/DNp01 `named` cut; the red line is the same contrast on the real cut. Columns run in increasing order of destruction: transmitter labels permuted while holding outgoing-weight shares near their real values, synapse weights permuted, degree-preserving double-edge rewiring, Erdős–Rényi. Insets show the central 99% of the concentrated Erdős–Rényi null distributions. Titles carry the empirical two-sided permutation p and the three-way verdict against the declared point-gap margin (delta = 0.332 model rate units, 5 % of the vehicle readout): `distinguishable` means the test rejected, `within point-gap tolerance` means the gap from the null median is below the margin after non-rejection, and `not resolved` means neither -- no label is a statistical equivalence test. The plain label permutation is documented in T23: it moves weighted excitation/inhibition balance and is a joint null rather than a ladder rung. Imidacloprid is composition-dominated (necessary level: degree_sequence, equivalent_within_tolerance); fipronil is topology-dependent (necessary level: wiring_with_weighted_transmitter_balance). -**F7_selectivity_landscape.** Each point is one compound on one committed graph. The x axis is the receptor selectivity index from the library alone; the y axis is the circuit index, the log10 width of the window between the concentration that moves the simulated network mean rate by 50% and the concentration at which the most potent vertebrate target reaches 20% occupancy. Points below the dashed identity line are buffered by the circuit, points above are amplified. The split is by mechanism, not by potency: chlorpyrifos_oxon, ddt, deltamethrin, permethrin (Nav modulators and the AChE inhibitor) amplify, mean gap +0.29 log10 units, while the nAChR-acting set (acetamiprid, acetylcholine, clothianidin, imidacloprid, nicotine, nitenpyram) buffers, mean gap -0.87. Compounds with no circuit index (chlordimeform, dieldrin, fipronil, gaba, ivermectin, picrotoxin, spinosad, sulfoxaflor, thiamethoxam) never reach the 50% threshold on this cut and are omitted. +**F7_selectivity_landscape.** Each point is one compound on one committed graph. The x axis is the receptor selectivity index from the library alone; the y axis is the circuit index, the log10 width of the window between the concentration that moves the simulated network mean rate by 50% and the concentration at which the most potent vertebrate target reaches 20% occupancy. Points below the dashed identity line are buffered by the circuit, points above are amplified. The split is by mechanism, not by potency: ddt, deltamethrin (Nav modulators and the AChE inhibitor) amplify, mean gap +0.19 log10 units, while the nAChR-acting set (acetamiprid, acetylcholine, clothianidin, imidacloprid, nicotine, nitenpyram) buffers, mean gap -0.87. Compounds with no circuit index (chlordimeform, dieldrin, fipronil, gaba, ivermectin, permethrin, picrotoxin, thiamethoxam) never reach the 50% threshold on this cut and are omitted. **F8_dose_response_sensitivity.** Left: the model's own dose-response for imidacloprid on the 1-hop MN9/DNp01 cut. Points are 4 replicates per concentration with log-normal EC50 jitter (sd 0.3 log10 units) and drive jitter; the curve is a 4-parameter Hill fit, and the band is a 200-resample bootstrap CI on the fitted IC50 (34.7 nM). This is the *model's* circuit IC50, not a receptor IC50 and not an animal IC50. Right: one-at-a-time sensitivity of the same readout at 1 µM, each parameter halved and doubled. The gain-rule coefficient dominates; the EC50 and Hill coefficient have zero span because occupancy is already saturated at this dose. -**F9_genotype_panel.** Target-receptor occupancy at 1 uM for wild type and each published resistance allele that has a sourced fold-shift for that compound. Grey bars are alleles with no sourced number for the compound: they are left identical to wild type and reported, never extrapolated. Left and centre: fipronil and imidacloprid. Right: the specificity check that motivates the per-compound rule. *para* M918T (super-kdr) shifts deltamethrin 100-fold but leaves DDT at 1-fold, and *Rdl* A301S shifts GABA 33.9-fold while leaving fipronil at 1.1-fold. A toggle that shifted a whole receptor uniformly would be wrong in both cases. +**F9_genotype_panel.** Model target engagement at 1 uM for wild type and each published resistance allele that has a sourced fold-shift for that compound. Grey bars are alleles with no sourced number for the compound: they are left identical to wild type and reported, never extrapolated. Left and centre: fipronil and imidacloprid. Right: the specificity check that motivates the per-compound rule. *para* M918T (super-kdr) shifts deltamethrin 100-fold but leaves DDT at 1-fold, and *Rdl* A301S shifts GABA 33.9-fold while leaving fipronil at 1.1-fold. A toggle that shifted a whole receptor uniformly would be wrong in both cases. diff --git a/papers/results.json b/papers/results.json index ede5eaf..426737f 100644 --- a/papers/results.json +++ b/papers/results.json @@ -1,18 +1,18 @@ { "schema": "flylab-paper-results/1", - "generated_utc": "2026-09-20T15:48:02+00:00", + "generated_utc": "2026-09-24T13:53:45+00:00", "fast": false, "seed": 0, "jobs": 4, - "runtime_s": 3824.39, + "runtime_s": 293.94, "provenance": { "flylab_version": "0.6.0", - "git_sha": "f95fd88a2165e66d5203db6c6fa8c6a020c5a8e0", - "library_sha256": "594b8d32c5e61ebd4de31cc77441aeb82503a0275a7a6ef951ec7a86a48732a8", + "git_sha": "50f5465b2df028a6cb2b93926cfbd14d7c746c9b", + "library_sha256": "359096467b33242174dcc5d787a6b7fc6d37573f8daf4771efd159b64ca95e05", "library_version": "0.6", "library_schema_version": 3, "notebook_schema_version": "0.3", - "platform": "Linux-6.18.44-fc-v37-x86_64-with-glibc2.39 python-3.11.15", + "platform": "macOS-26.7-x86_64-i386-64bit python-3.11.15", "map_id": "male-cns:v1.0", "map_citation": "MaleCNS v1.0, HHMI Janelia FlyEM / Cambridge / MRC LMB / Google Research. CC-BY 4.0.", "graphs": { @@ -47,7 +47,7 @@ }, "scorecard": { "status": "ok", - "runtime_s": 3.39, + "runtime_s": 0.57, "description": "insect vs vertebrate panel + library table (F2, T1)", "figures": 1, "tables": 2, @@ -55,7 +55,7 @@ }, "evidence": { "status": "ok", - "runtime_s": 2.41, + "runtime_s": 0.2, "description": "typed evidence census, schema v3 (F11, T10)", "figures": 1, "tables": 4, @@ -63,7 +63,7 @@ }, "curves": { "status": "ok", - "runtime_s": 6.43, + "runtime_s": 0.84, "description": "engagement curves and selectivity windows (F3)", "figures": 1, "tables": 0, @@ -71,7 +71,7 @@ }, "mechanisms": { "status": "ok", - "runtime_s": 0.01, + "runtime_s": 0.0, "description": "mechanism -> gain rules (T2)", "figures": 0, "tables": 4, @@ -79,7 +79,7 @@ }, "graph": { "status": "ok", - "runtime_s": 7.26, + "runtime_s": 0.82, "description": "taste-motor graph and GRN->MN9 paths (F4, T7)", "figures": 1, "tables": 4, @@ -87,7 +87,7 @@ }, "veto": { "status": "ok", - "runtime_s": 26.8, + "runtime_s": 3.81, "description": "map-based bitter veto, rate vs LIF (F5)", "figures": 1, "tables": 0, @@ -95,7 +95,7 @@ }, "dependence": { "status": "ok", - "runtime_s": 3610.41, + "runtime_s": 2262.02, "description": "connectome dependence + landscape (F6, F12, T6, T11, T12)", "figures": 3, "tables": 12, @@ -103,7 +103,7 @@ }, "ablation": { "status": "ok", - "runtime_s": 77.87, + "runtime_s": 32.94, "description": "ablation ladder over the library (F13, T13, T13b)", "figures": 1, "tables": 8, @@ -111,7 +111,7 @@ }, "landscape": { "status": "ok", - "runtime_s": 63.87, + "runtime_s": 45.26, "description": "receptor vs circuit selectivity (F7, T8)", "figures": 1, "tables": 2, @@ -119,7 +119,7 @@ }, "dose": { "status": "ok", - "runtime_s": 6.94, + "runtime_s": 3.73, "description": "model dose-response + sensitivity tornado (F8)", "figures": 1, "tables": 0, @@ -127,7 +127,7 @@ }, "genotype": { "status": "ok", - "runtime_s": 0.51, + "runtime_s": 0.58, "description": "resistance-allele panels (F9, T9)", "figures": 1, "tables": 2, @@ -135,7 +135,7 @@ }, "validation": { "status": "ok", - "runtime_s": 0.89, + "runtime_s": 0.55, "description": "rank comparison vs published orders (F10, T4)", "figures": 1, "tables": 2, @@ -143,7 +143,7 @@ }, "predictions": { "status": "ok", - "runtime_s": 15.08, + "runtime_s": 12.95, "description": "prospective predictions (T3)", "figures": 0, "tables": 2, @@ -151,47 +151,31 @@ }, "mixtures": { "status": "ok", - "runtime_s": 0.36, + "runtime_s": 0.31, "description": "binary mixtures vs published verdicts (T5)", "figures": 0, "tables": 2, "values": 6 }, - "expression": { - "status": "ok", - "runtime_s": 0.19, - "description": "expression-weighted sensitivity coverage", - "figures": 0, - "tables": 0, - "values": 14 - }, "stability": { "status": "ok", - "runtime_s": 0.62, + "runtime_s": 0.27, "description": "specification robustness + threshold grid (F14, T14, T14b, T15)", "figures": 1, "tables": 6, "values": 117 }, - "uncertainty": { - "status": "ok", - "runtime_s": 0.6, - "description": "global uncertainty budget + VOI (F15, T16, T17)", - "figures": 1, - "tables": 4, - "values": 76 - }, - "claims": { + "scale": { "status": "ok", - "runtime_s": 0.01, - "description": "claim provenance chain (T18)", + "runtime_s": 0.02, + "description": "where the dependence verdict settles with the cut (T26, T27 when the study exists)", "figures": 0, - "tables": 2, - "values": 15 + "tables": 4, + "values": 38 }, "architecture": { "status": "ok", - "runtime_s": 0.42, + "runtime_s": 0.19, "description": "architecture schematic (F1)", "figures": 2, "tables": 0, @@ -199,19 +183,35 @@ }, "provenance": { "status": "ok", - "runtime_s": 0.26, + "runtime_s": 0.34, "description": "versions, hashes, map ids, graph sizes", "figures": 0, "tables": 0, "values": 11 }, - "scale": { + "expression": { "status": "ok", - "runtime_s": 0.05, - "description": "where the dependence verdict settles with the cut (T26, T27 when the study exists)", + "runtime_s": 0.37, + "description": "expression-weighted sensitivity coverage", "figures": 0, + "tables": 0, + "values": 14 + }, + "uncertainty": { + "status": "ok", + "runtime_s": 293.14, + "description": "global uncertainty budget + VOI (F15, T16, T17)", + "figures": 1, "tables": 4, - "values": 38 + "values": 76 + }, + "claims": { + "status": "ok", + "runtime_s": 0.02, + "description": "claim provenance chain (T18)", + "figures": 0, + "tables": 2, + "values": 15 }, "paper": { "status": "ok", @@ -449,15 +449,15 @@ "step": "scorecard" }, "library_rows_sourced": { - "value": 46, - "text": "46", + "value": 39, + "text": "39", "unit": null, "note": null, "step": "scorecard" }, "library_rows_placeholder": { - "value": 56, - "text": "56", + "value": 63, + "text": "63", "unit": null, "note": null, "step": "scorecard" @@ -498,29 +498,29 @@ "step": "evidence" }, "ev_rows_modelled": { - "value": 46, - "text": "46", + "value": 39, + "text": "39", "unit": null, "note": null, "step": "evidence" }, "ev_rows_not_modelled": { - "value": 56, - "text": "56", + "value": 63, + "text": "63", "unit": null, "note": null, "step": "evidence" }, "ev_param_EC50": { - "value": 34, - "text": "34", + "value": 29, + "text": "29", "unit": null, "note": null, "step": "evidence" }, "ev_param_IC50": { - "value": 10, - "text": "10", + "value": 8, + "text": "8", "unit": null, "note": null, "step": "evidence" @@ -547,8 +547,8 @@ "step": "evidence" }, "ev_param_unknown": { - "value": 56, - "text": "56", + "value": 63, + "text": "63", "unit": null, "note": null, "step": "evidence" @@ -568,8 +568,8 @@ "step": "evidence" }, "ev_rel_exact_compound_exact_receptor_exact_species": { - "value": 10, - "text": "10", + "value": 7, + "text": "7", "unit": null, "note": null, "step": "evidence" @@ -582,22 +582,22 @@ "step": "evidence" }, "ev_rel_exact_compound_related_receptor": { - "value": 10, - "text": "10", + "value": 8, + "text": "8", "unit": null, "note": null, "step": "evidence" }, "ev_rel_class_extrapolation": { - "value": 11, - "text": "11", + "value": 9, + "text": "9", "unit": null, "note": null, "step": "evidence" }, "ev_rel_unsupported": { - "value": 56, - "text": "56", + "value": 63, + "text": "63", "unit": null, "note": null, "step": "evidence" @@ -617,22 +617,22 @@ "step": "evidence" }, "ev_model_functional_engagement": { - "value": 9, - "text": "9", + "value": 6, + "text": "6", "unit": null, "note": null, "step": "evidence" }, "ev_model_functional_engagement_proxy": { - "value": 35, - "text": "35", + "value": 31, + "text": "31", "unit": null, "note": null, "step": "evidence" }, "ev_model_not_modelled": { - "value": 56, - "text": "56", + "value": 63, + "text": "63", "unit": null, "note": null, "step": "evidence" @@ -645,29 +645,29 @@ "step": "evidence" }, "ev_functional_rows": { - "value": 9, - "text": "9", + "value": 6, + "text": "6", "unit": null, "note": null, "step": "evidence" }, "ev_not_modelled_rows": { - "value": 56, - "text": "56", + "value": 63, + "text": "63", "unit": null, "note": null, "step": "evidence" }, "ev_proxy_rows": { - "value": 36, - "text": "36", + "value": 32, + "text": "32", "unit": null, "note": null, "step": "evidence" }, "ev_dist_E0": { - "value": 10, - "text": "10", + "value": 7, + "text": "7", "unit": null, "note": null, "step": "evidence" @@ -680,22 +680,22 @@ "step": "evidence" }, "ev_dist_E2": { - "value": 10, - "text": "10", + "value": 8, + "text": "8", "unit": null, "note": null, "step": "evidence" }, "ev_dist_E3": { - "value": 11, - "text": "11", + "value": 9, + "text": "9", "unit": null, "note": null, "step": "evidence" }, "ev_dist_E4": { - "value": 56, - "text": "56", + "value": 63, + "text": "63", "unit": null, "note": null, "step": "evidence" @@ -781,7 +781,7 @@ "step": "curves" }, "window_log10_deltamethrin": { - "value": 2.0655015487564325, + "value": 2.065501548756432, "text": "2.07", "unit": null, "note": null, @@ -1247,16 +1247,16 @@ "step": "graph" }, "veto_rate_vehicle_sugar": { - "value": 0.2931028813192072, + "value": 0.29310288131920725, "text": "0.293", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "veto" }, "veto_rate_vehicle_sugar_bitter": { "value": 0.0, "text": "0.000", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "veto" }, @@ -1270,19 +1270,19 @@ "veto_rate_fipronil_sugar": { "value": 4.177454859032482, "text": "4.177", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "veto" }, "veto_rate_fipronil_sugar_bitter": { - "value": 1.7236839300867062, + "value": 1.723683930086706, "text": "1.724", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "veto" }, "veto_rate_fipronil_ratio": { - "value": 0.412615812319255, + "value": 0.4126158123192549, "text": "0.413", "unit": null, "note": null, @@ -1291,14 +1291,14 @@ "veto_rate_imidacloprid_sugar": { "value": 0.0, "text": "0.000", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "veto" }, "veto_rate_imidacloprid_sugar_bitter": { "value": 0.0, "text": "0.000", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "veto" }, @@ -1611,7 +1611,7 @@ "dep_named_imidacloprid_effect": { "value": -6.17278216613422, "text": "-6.173", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dependence" }, @@ -1646,7 +1646,7 @@ "dep_named_imidacloprid_delta": { "value": 0.33156221429633564, "text": "0.332", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dependence" }, @@ -2016,7 +2016,7 @@ "dep_named_fipronil_effect": { "value": 0.8996076235547861, "text": "0.900", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dependence" }, @@ -2051,7 +2051,7 @@ "dep_named_fipronil_delta": { "value": 0.33156221429633564, "text": "0.332", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dependence" }, @@ -2454,7 +2454,7 @@ "dep_taste_fipronil_delta": { "value": null, "text": "n/a", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dependence" }, @@ -3055,22 +3055,22 @@ "step": "dependence" }, "dep_land_topology_dependent": { - "value": 11, - "text": "11", + "value": 0, + "text": "0", "unit": null, "note": null, "step": "dependence" }, "dep_land_topology_dependent_raw": { - "value": 14, - "text": "14", + "value": 11, + "text": "11", "unit": null, "note": null, "step": "dependence" }, "dep_land_topology_dependent_absfloor": { - "value": 16, - "text": "16", + "value": 0, + "text": "0", "unit": null, "note": null, "step": "dependence" @@ -3097,43 +3097,43 @@ "step": "dependence" }, "dep_land_composition_dominated": { - "value": 42, - "text": "42", + "value": 44, + "text": "44", "unit": null, "note": null, "step": "dependence" }, "dep_land_composition_dominated_raw": { - "value": 39, - "text": "39", + "value": 33, + "text": "33", "unit": null, "note": null, "step": "dependence" }, "dep_land_composition_dominated_absfloor": { - "value": 55, - "text": "55", + "value": 59, + "text": "59", "unit": null, "note": null, "step": "dependence" }, "dep_land_no_effect": { - "value": 31, - "text": "31", + "value": 40, + "text": "40", "unit": null, "note": null, "step": "dependence" }, "dep_land_no_effect_raw": { - "value": 31, - "text": "31", + "value": 40, + "text": "40", "unit": null, "note": null, "step": "dependence" }, "dep_land_no_effect_absfloor": { - "value": 13, - "text": "13", + "value": 25, + "text": "25", "unit": null, "note": null, "step": "dependence" @@ -3167,8 +3167,8 @@ "step": "dependence" }, "dep_land_statement": { - "value": "exploratory screen, FDR-controlled: 11 of 84 cells are topology-dependent after Benjamini-Hochberg at fdr_alpha = 0.05 across 168 structural tests (14 before correction).", - "text": "exploratory screen, FDR-controlled: 11 of 84 cells are topology-dependent after Benjamini-Hochberg at fdr_alpha = 0.05 across 168 structural tests (14 before correction).", + "value": "exploratory screen with BH-adjusted structural component tests: 0 of 84 cells are topology-dependent after Benjamini-Hochberg at fdr_alpha = 0.05 across 168 structural tests (11 before correction). The derived cell classes have no separate cell-level FDR guarantee.", + "text": "exploratory screen with BH-adjusted structural component tests: 0 of 84 cells are topology-dependent after Benjamini-Hochberg at fdr_alpha = 0.05 across 168 structural tests (11 before correction). The derived cell classes have no separate cell-level FDR guarantee.", "unit": null, "note": null, "step": "dependence" @@ -3188,8 +3188,8 @@ "step": "dependence" }, "dep_land_n_structural_rejected": { - "value": 17, - "text": "17", + "value": 0, + "text": "0", "unit": null, "note": null, "step": "dependence" @@ -3202,15 +3202,15 @@ "step": "dependence" }, "dep_land_non_monotone": { - "value": 11, - "text": "11", + "value": 0, + "text": "0", "unit": null, "note": null, "step": "dependence" }, "dep_land_below_relative_floor": { - "value": 18, - "text": "18", + "value": 15, + "text": "15", "unit": null, "note": null, "step": "dependence" @@ -3226,102 +3226,67 @@ "gaba@1e-05", "ivermectin@1e-08", "ivermectin@1e-07", + "ivermectin@1e-06", + "ivermectin@1e-05", "picrotoxin@1e-08", - "spinosad@1e-08", - "spinosad@1e-07", - "sulfoxaflor@1e-08", - "sulfoxaflor@1e-07", - "sulfoxaflor@1e-06", "thiamethoxam@1e-08", "thiamethoxam@1e-07", "thiamethoxam@1e-06" ], - "text": "acetylcholine 10 nM (0.000761 Hz), acetylcholine 100 nM (0.0118 Hz), ddt 10 nM (0.00309 Hz), ddt 100 nM (0.0309 Hz), gaba 100 nM (-2.74e-05 Hz), gaba 1 \u00b5M (-0.000864 Hz), gaba 10 \u00b5M (-0.025 Hz), ivermectin 10 nM (-0.0156 Hz), ivermectin 100 nM (-0.0587 Hz), picrotoxin 10 nM (0.0076 Hz), spinosad 10 nM (0.00601 Hz), spinosad 100 nM (0.0549 Hz), sulfoxaflor 10 nM (0.000202 Hz), sulfoxaflor 100 nM (0.00202 Hz), sulfoxaflor 1 \u00b5M (0.0196 Hz), thiamethoxam 10 nM (0.000303 Hz), thiamethoxam 100 nM (0.00302 Hz), thiamethoxam 1 \u00b5M (0.0289 Hz)", + "text": "acetylcholine 10 nM (0.000761 rate-model units), acetylcholine 100 nM (0.0118 rate-model units), ddt 10 nM (0.00309 rate-model units), ddt 100 nM (0.0309 rate-model units), gaba 100 nM (-2.74e-05 rate-model units), gaba 1 \u00b5M (-0.000864 rate-model units), gaba 10 \u00b5M (-0.025 rate-model units), ivermectin 10 nM (-0.0125 rate-model units), ivermectin 100 nM (-0.0312 rate-model units), ivermectin 1 \u00b5M (-0.0367 rate-model units), ivermectin 10 \u00b5M (-0.0373 rate-model units), picrotoxin 10 nM (0.0076 rate-model units), thiamethoxam 10 nM (0.000303 rate-model units), thiamethoxam 100 nM (0.00302 rate-model units), thiamethoxam 1 \u00b5M (0.0289 rate-model units)", "unit": null, "note": null, "step": "dependence" }, "dep_land_non_monotone_cells": { - "value": [ - "chlorpyrifos_oxon@1e-08", - "dieldrin@1e-08", - "dieldrin@1e-07", - "dieldrin@1e-06", - "fipronil@1e-08", - "fipronil@1e-07", - "ivermectin@1e-06", - "ivermectin@1e-05", - "picrotoxin@1e-07", - "picrotoxin@1e-06", - "picrotoxin@1e-05" - ], - "text": "chlorpyrifos_oxon 10 nM, dieldrin 10 nM, dieldrin 100 nM, dieldrin 1 \u00b5M, fipronil 10 nM, fipronil 100 nM, ivermectin 1 \u00b5M, ivermectin 10 \u00b5M, picrotoxin 100 nM, picrotoxin 1 \u00b5M, picrotoxin 10 \u00b5M", + "value": [], + "text": "none", "unit": null, "note": null, "step": "dependence" }, "dep_land_topology_compounds": { - "value": [ - "chlorpyrifos_oxon", - "dieldrin", - "fipronil", - "ivermectin", - "picrotoxin" - ], - "text": "chlorpyrifos_oxon, dieldrin, fipronil, ivermectin, picrotoxin", + "value": [], + "text": "none", "unit": null, "note": null, "step": "dependence" }, "dep_land_topology_compounds_raw": { "value": [ - "chlorpyrifos_oxon", "dieldrin", "fipronil", - "ivermectin", "picrotoxin" ], - "text": "chlorpyrifos_oxon, dieldrin, fipronil, ivermectin, picrotoxin", + "text": "dieldrin, fipronil, picrotoxin", "unit": null, "note": null, "step": "dependence" }, "dep_land_topology_cells": { - "value": [ - "chlorpyrifos_oxon@1e-08", - "dieldrin@1e-08", - "dieldrin@1e-07", - "dieldrin@1e-06", - "fipronil@1e-08", - "fipronil@1e-07", - "ivermectin@1e-06", - "ivermectin@1e-05", - "picrotoxin@1e-07", - "picrotoxin@1e-06", - "picrotoxin@1e-05" - ], - "text": "chlorpyrifos_oxon 10 nM, dieldrin 10 nM, dieldrin 100 nM, dieldrin 1 \u00b5M, fipronil 10 nM, fipronil 100 nM, ivermectin 1 \u00b5M, ivermectin 10 \u00b5M, picrotoxin 100 nM, picrotoxin 1 \u00b5M, picrotoxin 10 \u00b5M", + "value": [], + "text": "none", "unit": null, "note": null, "step": "dependence" }, "dep_land_verdicts_distinguishable": { - "value": 89, - "text": "89", + "value": 60, + "text": "60", "unit": null, "note": null, "step": "dependence" }, "dep_land_verdicts_equivalent_within_tolerance": { - "value": 281, - "text": "281", + "value": 316, + "text": "316", "unit": null, "note": null, "step": "dependence" }, "dep_land_verdicts_indeterminate": { - "value": 50, - "text": "50", + "value": 44, + "text": "44", "unit": null, "note": null, "step": "dependence" @@ -3329,14 +3294,14 @@ "dep_land_delta_min": { "value": 0.33156221429633564, "text": "0.332", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dependence" }, "dep_land_delta_max": { "value": 0.33156221429633564, "text": "0.332", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dependence" }, @@ -3355,8 +3320,8 @@ "step": "dependence" }, "dep_land_fipronil_q": { - "value": 0.05034965034965035, - "text": "0.050", + "value": 0.061832903938167094, + "text": "0.062", "unit": null, "note": null, "step": "dependence" @@ -3418,8 +3383,8 @@ "step": "dependence" }, "dep_land_imidacloprid_q": { - "value": 0.9038301148392892, - "text": "0.904", + "value": 1.0, + "text": "1.000", "unit": null, "note": null, "step": "dependence" @@ -3495,22 +3460,22 @@ "step": "dependence" }, "dep_land_taste_topology_dependent": { - "value": 54, - "text": "54", + "value": 44, + "text": "44", "unit": null, "note": null, "step": "dependence" }, "dep_land_taste_topology_dependent_raw": { - "value": 54, - "text": "54", + "value": 44, + "text": "44", "unit": null, "note": null, "step": "dependence" }, "dep_land_taste_topology_dependent_absfloor": { - "value": 71, - "text": "71", + "value": 60, + "text": "60", "unit": null, "note": null, "step": "dependence" @@ -3551,29 +3516,29 @@ "step": "dependence" }, "dep_land_taste_composition_dominated_absfloor": { - "value": 1, - "text": "1", + "value": 0, + "text": "0", "unit": null, "note": null, "step": "dependence" }, "dep_land_taste_no_effect": { - "value": 30, - "text": "30", + "value": 40, + "text": "40", "unit": null, "note": null, "step": "dependence" }, "dep_land_taste_no_effect_raw": { - "value": 30, - "text": "30", + "value": 40, + "text": "40", "unit": null, "note": null, "step": "dependence" }, "dep_land_taste_no_effect_absfloor": { - "value": 12, - "text": "12", + "value": 24, + "text": "24", "unit": null, "note": null, "step": "dependence" @@ -3607,8 +3572,8 @@ "step": "dependence" }, "dep_land_taste_statement": { - "value": "exploratory screen, FDR-controlled: 54 of 84 cells are topology-dependent after Benjamini-Hochberg at fdr_alpha = 0.05 across 168 structural tests (54 before correction).", - "text": "exploratory screen, FDR-controlled: 54 of 84 cells are topology-dependent after Benjamini-Hochberg at fdr_alpha = 0.05 across 168 structural tests (54 before correction).", + "value": "exploratory screen with BH-adjusted structural component tests: 44 of 84 cells are topology-dependent after Benjamini-Hochberg at fdr_alpha = 0.05 across 168 structural tests (44 before correction). The derived cell classes have no separate cell-level FDR guarantee.", + "text": "exploratory screen with BH-adjusted structural component tests: 44 of 84 cells are topology-dependent after Benjamini-Hochberg at fdr_alpha = 0.05 across 168 structural tests (44 before correction). The derived cell classes have no separate cell-level FDR guarantee.", "unit": null, "note": null, "step": "dependence" @@ -3628,8 +3593,8 @@ "step": "dependence" }, "dep_land_taste_n_structural_rejected": { - "value": 92, - "text": "92", + "value": 77, + "text": "77", "unit": null, "note": null, "step": "dependence" @@ -3642,15 +3607,15 @@ "step": "dependence" }, "dep_land_taste_non_monotone": { - "value": 0, - "text": "0", + "value": 1, + "text": "1", "unit": null, "note": null, "step": "dependence" }, "dep_land_taste_below_relative_floor": { - "value": 18, - "text": "18", + "value": 16, + "text": "16", "unit": null, "note": null, "step": "dependence" @@ -3666,24 +3631,24 @@ "gaba@1e-06", "gaba@1e-05", "ivermectin@1e-08", + "ivermectin@1e-07", + "ivermectin@1e-06", + "ivermectin@1e-05", "picrotoxin@1e-08", - "spinosad@1e-08", - "spinosad@1e-07", - "sulfoxaflor@1e-08", - "sulfoxaflor@1e-07", - "sulfoxaflor@1e-06", "thiamethoxam@1e-08", "thiamethoxam@1e-07", "thiamethoxam@1e-06" ], - "text": "acetylcholine 10 nM (0.0018 Hz), acetylcholine 100 nM (0.028 Hz), ddt 10 nM (0.00656 Hz), ddt 100 nM (0.0656 Hz), gaba 10 nM (-4.27e-06 Hz), gaba 100 nM (-0.000135 Hz), gaba 1 \u00b5M (-0.00426 Hz), gaba 10 \u00b5M (-0.122 Hz), ivermectin 10 nM (-0.0348 Hz), picrotoxin 10 nM (0.0375 Hz), spinosad 10 nM (0.0142 Hz), spinosad 100 nM (0.13 Hz), sulfoxaflor 10 nM (0.000478 Hz), sulfoxaflor 100 nM (0.00477 Hz), sulfoxaflor 1 \u00b5M (0.0463 Hz), thiamethoxam 10 nM (0.000717 Hz), thiamethoxam 100 nM (0.00714 Hz), thiamethoxam 1 \u00b5M (0.0683 Hz)", + "text": "acetylcholine 10 nM (0.0018 rate-model units), acetylcholine 100 nM (0.028 rate-model units), ddt 10 nM (0.00656 rate-model units), ddt 100 nM (0.0656 rate-model units), gaba 10 nM (-4.27e-06 rate-model units), gaba 100 nM (-0.000135 rate-model units), gaba 1 \u00b5M (-0.00426 rate-model units), gaba 10 \u00b5M (-0.122 rate-model units), ivermectin 10 nM (-0.02 rate-model units), ivermectin 100 nM (-0.0454 rate-model units), ivermectin 1 \u00b5M (-0.0523 rate-model units), ivermectin 10 \u00b5M (-0.0532 rate-model units), picrotoxin 10 nM (0.0375 rate-model units), thiamethoxam 10 nM (0.000717 rate-model units), thiamethoxam 100 nM (0.00714 rate-model units), thiamethoxam 1 \u00b5M (0.0683 rate-model units)", "unit": null, "note": null, "step": "dependence" }, "dep_land_taste_non_monotone_cells": { - "value": [], - "text": "none", + "value": [ + "deltamethrin@1e-07" + ], + "text": "deltamethrin 100 nM", "unit": null, "note": null, "step": "dependence" @@ -3692,23 +3657,19 @@ "value": [ "acetamiprid", "acetylcholine", - "chlorpyrifos_oxon", "clothianidin", "ddt", "deltamethrin", "dieldrin", "fipronil", "imidacloprid", - "ivermectin", "nicotine", "nitenpyram", "permethrin", "picrotoxin", - "spinosad", - "sulfoxaflor", "thiamethoxam" ], - "text": "acetamiprid, acetylcholine, chlorpyrifos_oxon, clothianidin, ddt, deltamethrin, dieldrin, fipronil, imidacloprid, ivermectin, nicotine, nitenpyram, permethrin, picrotoxin, spinosad, sulfoxaflor, thiamethoxam", + "text": "acetamiprid, acetylcholine, clothianidin, ddt, deltamethrin, dieldrin, fipronil, imidacloprid, nicotine, nitenpyram, permethrin, picrotoxin, thiamethoxam", "unit": null, "note": null, "step": "dependence" @@ -3717,23 +3678,19 @@ "value": [ "acetamiprid", "acetylcholine", - "chlorpyrifos_oxon", "clothianidin", "ddt", "deltamethrin", "dieldrin", "fipronil", "imidacloprid", - "ivermectin", "nicotine", "nitenpyram", "permethrin", "picrotoxin", - "spinosad", - "sulfoxaflor", "thiamethoxam" ], - "text": "acetamiprid, acetylcholine, chlorpyrifos_oxon, clothianidin, ddt, deltamethrin, dieldrin, fipronil, imidacloprid, ivermectin, nicotine, nitenpyram, permethrin, picrotoxin, spinosad, sulfoxaflor, thiamethoxam", + "text": "acetamiprid, acetylcholine, clothianidin, ddt, deltamethrin, dieldrin, fipronil, imidacloprid, nicotine, nitenpyram, permethrin, picrotoxin, thiamethoxam", "unit": null, "note": null, "step": "dependence" @@ -3746,10 +3703,6 @@ "acetamiprid@1e-05", "acetylcholine@1e-06", "acetylcholine@1e-05", - "chlorpyrifos_oxon@1e-08", - "chlorpyrifos_oxon@1e-07", - "chlorpyrifos_oxon@1e-06", - "chlorpyrifos_oxon@1e-05", "clothianidin@1e-08", "clothianidin@1e-07", "clothianidin@1e-06", @@ -3772,9 +3725,6 @@ "imidacloprid@1e-07", "imidacloprid@1e-06", "imidacloprid@1e-05", - "ivermectin@1e-07", - "ivermectin@1e-06", - "ivermectin@1e-05", "nicotine@1e-08", "nicotine@1e-07", "nicotine@1e-06", @@ -3790,33 +3740,30 @@ "picrotoxin@1e-07", "picrotoxin@1e-06", "picrotoxin@1e-05", - "spinosad@1e-06", - "spinosad@1e-05", - "sulfoxaflor@1e-05", "thiamethoxam@1e-05" ], - "text": "acetamiprid 10 nM, acetamiprid 100 nM, acetamiprid 1 \u00b5M, acetamiprid 10 \u00b5M, acetylcholine 1 \u00b5M, acetylcholine 10 \u00b5M, chlorpyrifos_oxon 10 nM, chlorpyrifos_oxon 100 nM, chlorpyrifos_oxon 1 \u00b5M, chlorpyrifos_oxon 10 \u00b5M, clothianidin 10 nM, clothianidin 100 nM, clothianidin 1 \u00b5M, clothianidin 10 \u00b5M, ddt 1 \u00b5M, ddt 10 \u00b5M, deltamethrin 10 nM, deltamethrin 100 nM, deltamethrin 1 \u00b5M, deltamethrin 10 \u00b5M, dieldrin 10 nM, dieldrin 100 nM, dieldrin 1 \u00b5M, dieldrin 10 \u00b5M, fipronil 10 nM, fipronil 100 nM, fipronil 1 \u00b5M, fipronil 10 \u00b5M, imidacloprid 10 nM, imidacloprid 100 nM, imidacloprid 1 \u00b5M, imidacloprid 10 \u00b5M, ivermectin 100 nM, ivermectin 1 \u00b5M, ivermectin 10 \u00b5M, nicotine 10 nM, nicotine 100 nM, nicotine 1 \u00b5M, nicotine 10 \u00b5M, nitenpyram 10 nM, nitenpyram 100 nM, nitenpyram 1 \u00b5M, nitenpyram 10 \u00b5M, permethrin 10 nM, permethrin 100 nM, permethrin 1 \u00b5M, permethrin 10 \u00b5M, picrotoxin 100 nM, picrotoxin 1 \u00b5M, picrotoxin 10 \u00b5M, spinosad 1 \u00b5M, spinosad 10 \u00b5M, sulfoxaflor 10 \u00b5M, thiamethoxam 10 \u00b5M", + "text": "acetamiprid 10 nM, acetamiprid 100 nM, acetamiprid 1 \u00b5M, acetamiprid 10 \u00b5M, acetylcholine 1 \u00b5M, acetylcholine 10 \u00b5M, clothianidin 10 nM, clothianidin 100 nM, clothianidin 1 \u00b5M, clothianidin 10 \u00b5M, ddt 1 \u00b5M, ddt 10 \u00b5M, deltamethrin 10 nM, deltamethrin 100 nM, deltamethrin 1 \u00b5M, deltamethrin 10 \u00b5M, dieldrin 10 nM, dieldrin 100 nM, dieldrin 1 \u00b5M, dieldrin 10 \u00b5M, fipronil 10 nM, fipronil 100 nM, fipronil 1 \u00b5M, fipronil 10 \u00b5M, imidacloprid 10 nM, imidacloprid 100 nM, imidacloprid 1 \u00b5M, imidacloprid 10 \u00b5M, nicotine 10 nM, nicotine 100 nM, nicotine 1 \u00b5M, nicotine 10 \u00b5M, nitenpyram 10 nM, nitenpyram 100 nM, nitenpyram 1 \u00b5M, nitenpyram 10 \u00b5M, permethrin 10 nM, permethrin 100 nM, permethrin 1 \u00b5M, permethrin 10 \u00b5M, picrotoxin 100 nM, picrotoxin 1 \u00b5M, picrotoxin 10 \u00b5M, thiamethoxam 10 \u00b5M", "unit": null, "note": null, "step": "dependence" }, "dep_land_taste_verdicts_distinguishable": { - "value": 164, - "text": "164", + "value": 137, + "text": "137", "unit": null, "note": null, "step": "dependence" }, "dep_land_taste_verdicts_equivalent_within_tolerance": { - "value": 189, - "text": "189", + "value": 225, + "text": "225", "unit": null, "note": null, "step": "dependence" }, "dep_land_taste_verdicts_indeterminate": { - "value": 67, - "text": "67", + "value": 58, + "text": "58", "unit": null, "note": null, "step": "dependence" @@ -3824,14 +3771,14 @@ "dep_land_taste_delta_min": { "value": 0.7172853704242903, "text": "0.717", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dependence" }, "dep_land_taste_delta_max": { "value": 0.7172853704242903, "text": "0.717", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dependence" }, @@ -3850,8 +3797,8 @@ "step": "dependence" }, "dep_land_taste_fipronil_q": { - "value": 0.006489994591671174, - "text": "0.006", + "value": 0.008088978766430738, + "text": "0.008", "unit": null, "note": null, "step": "dependence" @@ -3913,8 +3860,8 @@ "step": "dependence" }, "dep_land_taste_imidacloprid_q": { - "value": 0.006489994591671174, - "text": "0.006", + "value": 0.008088978766430738, + "text": "0.008", "unit": null, "note": null, "step": "dependence" @@ -4023,49 +3970,49 @@ "step": "ablation" }, "abl_rho_receptor_min": { - "value": -0.5057464187739324, - "text": "-0.506", + "value": -0.5787938915672565, + "text": "-0.579", "unit": null, "note": null, "step": "ablation" }, "abl_rho_receptor_max": { - "value": -0.04620893838817564, - "text": "-0.046", + "value": 0.09527077257487362, + "text": "0.095", "unit": null, "note": null, "step": "ablation" }, "abl_rho_composition_min": { - "value": 0.9061889250814332, - "text": "0.906", + "value": 0.8763297872340425, + "text": "0.876", "unit": null, "note": null, "step": "ablation" }, "abl_rho_composition_max": { - "value": 0.9885475201437236, - "text": "0.989", + "value": 0.9541778975741239, + "text": "0.954", "unit": null, "note": null, "step": "ablation" }, "abl_rho_topology_min": { - "value": 0.9283856134172466, - "text": "0.928", + "value": 0.923901376977036, + "text": "0.924", "unit": null, "note": null, "step": "ablation" }, "abl_rho_topology_max": { - "value": 0.9625777447339687, - "text": "0.963", + "value": 0.9676140362132637, + "text": "0.968", "unit": null, "note": null, "step": "ablation" }, "abl_rho_full_flylab_min": { - "value": 1.0, + "value": 0.9999999999999999, "text": "1.000", "unit": null, "note": null, @@ -4079,66 +4026,66 @@ "step": "ablation" }, "abl_rho_topology_paper": { - "value": 0.9610814391588707, - "text": "0.961", + "value": 0.9676140362132637, + "text": "0.968", "unit": null, "note": null, "step": "ablation" }, "abl_rms_topology_hz": { - "value": 8.149381803321555, - "text": "8.149", - "unit": "Hz", + "value": 8.150695665444447, + "text": "8.151", + "unit": "rate-model units", "note": null, "step": "ablation" }, "abl_rho_composition_paper": { - "value": 0.9878103669773378, - "text": "0.988", + "value": 0.9214816983710898, + "text": "0.921", "unit": null, "note": null, "step": "ablation" }, "abl_rho_receptor_paper": { - "value": -0.4314654500371033, - "text": "-0.431", + "value": -0.5640523883085501, + "text": "-0.564", "unit": null, "note": null, "step": "ablation" }, "abl_rho_topology_floor_paper": { - "value": 0.24159822391955538, - "text": "0.242", + "value": 0.16087398593797766, + "text": "0.161", "unit": null, "note": null, "step": "ablation" }, "abl_statements": { "value": [ - "A_receptor_only does NOT reproduce the full model at 1.0e-08 M: rho = -0.046, r = -0.289; the layers above it are doing real work.", - "B_composition_only REPRODUCES the full model at 1.0e-08 M (ordering only: this level's units are not the full model's): rho = 0.906, r = 0.998 over 21 compounds. The layers above it add no ordering information for this readout.", - "C_topology_only reproduces the full model's ORDERING but not its values at 1.0e-08 M: rho = 0.928 (>= 0.9) with r = 0.470 (< 0.95) over 21 compounds.", - "C_topology_only at 1.0e-08 M: the direction-aware generic rule reaches rho = 0.928 against rho = -0.427 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +1.355, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome.", - "B_composition_only at 1.0e-08 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.906 to 0.964 (+0.057).", - "A_receptor_only does NOT reproduce the full model at 1.0e-07 M: rho = -0.348, r = -0.064; the layers above it are doing real work.", - "B_composition_only REPRODUCES the full model at 1.0e-07 M (ordering only: this level's units are not the full model's): rho = 0.989, r = 0.717 over 21 compounds. The layers above it add no ordering information for this readout.", - "C_topology_only reproduces the full model's ORDERING but not its values at 1.0e-07 M: rho = 0.963 (>= 0.9) with r = 0.638 (< 0.95) over 21 compounds.", - "C_topology_only at 1.0e-07 M: the direction-aware generic rule reaches rho = 0.963 against rho = 0.279 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +0.684, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome.", - "B_composition_only at 1.0e-07 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.989 to 0.987 (-0.001).", - "A_receptor_only does NOT reproduce the full model at 1.0e-06 M: rho = -0.431, r = 0.031; the layers above it are doing real work.", - "B_composition_only REPRODUCES the full model at 1.0e-06 M (ordering only: this level's units are not the full model's): rho = 0.988, r = 0.756 over 21 compounds. The layers above it add no ordering information for this readout.", - "C_topology_only reproduces the full model's ORDERING but not its values at 1.0e-06 M: rho = 0.961 (>= 0.9) with r = 0.654 (< 0.95) over 21 compounds.", - "C_topology_only at 1.0e-06 M: the direction-aware generic rule reaches rho = 0.961 against rho = 0.242 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +0.719, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome.", - "B_composition_only at 1.0e-06 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.988 to 0.984 (-0.004).", - "A_receptor_only does NOT reproduce the full model at 1.0e-05 M: rho = -0.506, r = 0.044; the layers above it are doing real work.", - "B_composition_only REPRODUCES the full model at 1.0e-05 M (ordering only: this level's units are not the full model's): rho = 0.987, r = 0.754 over 21 compounds. The layers above it add no ordering information for this readout.", - "C_topology_only reproduces the full model's ORDERING but not its values at 1.0e-05 M: rho = 0.939 (>= 0.9) with r = 0.651 (< 0.95) over 21 compounds.", - "C_topology_only at 1.0e-05 M: the direction-aware generic rule reaches rho = 0.939 against rho = 0.236 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +0.703, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome.", - "B_composition_only at 1.0e-05 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.987 to 0.979 (-0.008).", - "Concentration dependence of the ladder: the weakest level is A_receptor_only at every concentration tested (1e-08 M: A_receptor_only (-0.046), 1e-07 M: A_receptor_only (-0.348), 1e-06 M: A_receptor_only (-0.431), 1e-05 M: A_receptor_only (-0.506)).", + "A_receptor_only is below the declared rank threshold at 1.0e-08 M: rho = 0.095, r = -0.033.", + "B_composition_only is below the declared rank threshold at 1.0e-08 M: rho = 0.876, r = 0.760.", + "C_topology_only meets the rank threshold but not the value-agreement threshold at 1.0e-08 M: rho = 0.924 (>= 0.9) with r = 0.268 (< 0.95) over 21 compounds.", + "C_topology_only at 1.0e-08 M: the direction-aware generic rule reaches rho = 0.924 against rho = -0.371 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +1.295, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome.", + "B_composition_only at 1.0e-08 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.876 to 0.963 (+0.086).", + "A_receptor_only is below the declared rank threshold at 1.0e-07 M: rho = -0.519, r = -0.136.", + "B_composition_only meets the declared rank-agreement threshold at 1.0e-07 M (ordering only: this level's units are not the full model's): rho = 0.918, r = 0.706 over 21 compounds. This is agreement in the tested library, not independent validation of either model.", + "C_topology_only meets the rank threshold but not the value-agreement threshold at 1.0e-07 M: rho = 0.962 (>= 0.9) with r = 0.620 (< 0.95) over 21 compounds.", + "C_topology_only at 1.0e-07 M: the direction-aware generic rule reaches rho = 0.962 against rho = 0.175 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +0.787, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome.", + "B_composition_only at 1.0e-07 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.918 to 0.992 (+0.073).", + "A_receptor_only is below the declared rank threshold at 1.0e-06 M: rho = -0.564, r = -0.008.", + "B_composition_only meets the declared rank-agreement threshold at 1.0e-06 M (ordering only: this level's units are not the full model's): rho = 0.921, r = 0.750 over 21 compounds. This is agreement in the tested library, not independent validation of either model.", + "C_topology_only meets the rank threshold but not the value-agreement threshold at 1.0e-06 M: rho = 0.968 (>= 0.9) with r = 0.645 (< 0.95) over 21 compounds.", + "C_topology_only at 1.0e-06 M: the direction-aware generic rule reaches rho = 0.968 against rho = 0.161 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +0.807, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome.", + "B_composition_only at 1.0e-06 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.921 to 0.991 (+0.070).", + "A_receptor_only is below the declared rank threshold at 1.0e-05 M: rho = -0.579, r = 0.042.", + "B_composition_only meets the declared rank-agreement threshold at 1.0e-05 M (ordering only: this level's units are not the full model's): rho = 0.954, r = 0.750 over 21 compounds. This is agreement in the tested library, not independent validation of either model.", + "C_topology_only meets the rank threshold but not the value-agreement threshold at 1.0e-05 M: rho = 0.950 (>= 0.9) with r = 0.636 (< 0.95) over 21 compounds.", + "C_topology_only at 1.0e-05 M: the direction-aware generic rule reaches rho = 0.950 against rho = 0.151 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +0.800, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome.", + "B_composition_only at 1.0e-05 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.954 to 0.992 (+0.038).", + "Concentration dependence of the ladder: the weakest level is A_receptor_only at every concentration tested (1e-08 M: A_receptor_only (0.095), 1e-07 M: A_receptor_only (-0.519), 1e-06 M: A_receptor_only (-0.564), 1e-05 M: A_receptor_only (-0.579)).", "With the historical depression-only level-C rule the weakest level is 1e-08 M: C_topology_only, 1e-07 M: A_receptor_only, 1e-06 M: A_receptor_only, 1e-05 M: A_receptor_only: that rule is the worst of the four only at the lowest concentration, and at the three higher ones the receptor-only level is worse. Any sentence of the form 'the topology-only level is the worst of the four' is therefore true of one column of this table and false of the other three." ], - "text": "A_receptor_only does NOT reproduce the full model at 1.0e-08 M: rho = -0.046, r = -0.289; the layers above it are doing real work. B_composition_only REPRODUCES the full model at 1.0e-08 M (ordering only: this level's units are not the full model's): rho = 0.906, r = 0.998 over 21 compounds. The layers above it add no ordering information for this readout. C_topology_only reproduces the full model's ORDERING but not its values at 1.0e-08 M: rho = 0.928 (>= 0.9) with r = 0.470 (< 0.95) over 21 compounds. C_topology_only at 1.0e-08 M: the direction-aware generic rule reaches rho = 0.928 against rho = -0.427 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +1.355, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome. B_composition_only at 1.0e-08 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.906 to 0.964 (+0.057). A_receptor_only does NOT reproduce the full model at 1.0e-07 M: rho = -0.348, r = -0.064; the layers above it are doing real work. B_composition_only REPRODUCES the full model at 1.0e-07 M (ordering only: this level's units are not the full model's): rho = 0.989, r = 0.717 over 21 compounds. The layers above it add no ordering information for this readout. C_topology_only reproduces the full model's ORDERING but not its values at 1.0e-07 M: rho = 0.963 (>= 0.9) with r = 0.638 (< 0.95) over 21 compounds. C_topology_only at 1.0e-07 M: the direction-aware generic rule reaches rho = 0.963 against rho = 0.279 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +0.684, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome. B_composition_only at 1.0e-07 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.989 to 0.987 (-0.001). A_receptor_only does NOT reproduce the full model at 1.0e-06 M: rho = -0.431, r = 0.031; the layers above it are doing real work. B_composition_only REPRODUCES the full model at 1.0e-06 M (ordering only: this level's units are not the full model's): rho = 0.988, r = 0.756 over 21 compounds. The layers above it add no ordering information for this readout. C_topology_only reproduces the full model's ORDERING but not its values at 1.0e-06 M: rho = 0.961 (>= 0.9) with r = 0.654 (< 0.95) over 21 compounds. C_topology_only at 1.0e-06 M: the direction-aware generic rule reaches rho = 0.961 against rho = 0.242 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +0.719, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome. B_composition_only at 1.0e-06 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.988 to 0.984 (-0.004). A_receptor_only does NOT reproduce the full model at 1.0e-05 M: rho = -0.506, r = 0.044; the layers above it are doing real work. B_composition_only REPRODUCES the full model at 1.0e-05 M (ordering only: this level's units are not the full model's): rho = 0.987, r = 0.754 over 21 compounds. The layers above it add no ordering information for this readout. C_topology_only reproduces the full model's ORDERING but not its values at 1.0e-05 M: rho = 0.939 (>= 0.9) with r = 0.651 (< 0.95) over 21 compounds. C_topology_only at 1.0e-05 M: the direction-aware generic rule reaches rho = 0.939 against rho = 0.236 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +0.703, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome. B_composition_only at 1.0e-05 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.987 to 0.979 (-0.008). Concentration dependence of the ladder: the weakest level is A_receptor_only at every concentration tested (1e-08 M: A_receptor_only (-0.046), 1e-07 M: A_receptor_only (-0.348), 1e-06 M: A_receptor_only (-0.431), 1e-05 M: A_receptor_only (-0.506)). With the historical depression-only level-C rule the weakest level is 1e-08 M: C_topology_only, 1e-07 M: A_receptor_only, 1e-06 M: A_receptor_only, 1e-05 M: A_receptor_only: that rule is the worst of the four only at the lowest concentration, and at the three higher ones the receptor-only level is worse. Any sentence of the form 'the topology-only level is the worst of the four' is therefore true of one column of this table and false of the other three.", + "text": "A_receptor_only is below the declared rank threshold at 1.0e-08 M: rho = 0.095, r = -0.033. B_composition_only is below the declared rank threshold at 1.0e-08 M: rho = 0.876, r = 0.760. C_topology_only meets the rank threshold but not the value-agreement threshold at 1.0e-08 M: rho = 0.924 (>= 0.9) with r = 0.268 (< 0.95) over 21 compounds. C_topology_only at 1.0e-08 M: the direction-aware generic rule reaches rho = 0.924 against rho = -0.371 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +1.295, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome. B_composition_only at 1.0e-08 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.876 to 0.963 (+0.086). A_receptor_only is below the declared rank threshold at 1.0e-07 M: rho = -0.519, r = -0.136. B_composition_only meets the declared rank-agreement threshold at 1.0e-07 M (ordering only: this level's units are not the full model's): rho = 0.918, r = 0.706 over 21 compounds. This is agreement in the tested library, not independent validation of either model. C_topology_only meets the rank threshold but not the value-agreement threshold at 1.0e-07 M: rho = 0.962 (>= 0.9) with r = 0.620 (< 0.95) over 21 compounds. C_topology_only at 1.0e-07 M: the direction-aware generic rule reaches rho = 0.962 against rho = 0.175 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +0.787, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome. B_composition_only at 1.0e-07 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.918 to 0.992 (+0.073). A_receptor_only is below the declared rank threshold at 1.0e-06 M: rho = -0.564, r = -0.008. B_composition_only meets the declared rank-agreement threshold at 1.0e-06 M (ordering only: this level's units are not the full model's): rho = 0.921, r = 0.750 over 21 compounds. This is agreement in the tested library, not independent validation of either model. C_topology_only meets the rank threshold but not the value-agreement threshold at 1.0e-06 M: rho = 0.968 (>= 0.9) with r = 0.645 (< 0.95) over 21 compounds. C_topology_only at 1.0e-06 M: the direction-aware generic rule reaches rho = 0.968 against rho = 0.161 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +0.807, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome. B_composition_only at 1.0e-06 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.921 to 0.991 (+0.070). A_receptor_only is below the declared rank threshold at 1.0e-05 M: rho = -0.579, r = 0.042. B_composition_only meets the declared rank-agreement threshold at 1.0e-05 M (ordering only: this level's units are not the full model's): rho = 0.954, r = 0.750 over 21 compounds. This is agreement in the tested library, not independent validation of either model. C_topology_only meets the rank threshold but not the value-agreement threshold at 1.0e-05 M: rho = 0.950 (>= 0.9) with r = 0.636 (< 0.95) over 21 compounds. C_topology_only at 1.0e-05 M: the direction-aware generic rule reaches rho = 0.950 against rho = 0.151 for the historical depression-only rule (a floor, which cannot express disinhibition at all); the difference, +0.800, is how much of level C's old score was an artefact of that rule rather than a fact about the connectome. B_composition_only at 1.0e-05 M: the composition index signs glutamate as partly excitatory while the rate engine signs it inhibitory; reconciling them moves the B-versus-D rank correlation from 0.954 to 0.992 (+0.038). Concentration dependence of the ladder: the weakest level is A_receptor_only at every concentration tested (1e-08 M: A_receptor_only (0.095), 1e-07 M: A_receptor_only (-0.519), 1e-06 M: A_receptor_only (-0.564), 1e-05 M: A_receptor_only (-0.579)). With the historical depression-only level-C rule the weakest level is 1e-08 M: C_topology_only, 1e-07 M: A_receptor_only, 1e-06 M: A_receptor_only, 1e-05 M: A_receptor_only: that rule is the worst of the four only at the lowest concentration, and at the three higher ones the receptor-only level is worse. Any sentence of the form 'the topology-only level is the worst of the four' is therefore true of one column of this table and false of the other three.", "unit": null, "note": null, "step": "ablation" @@ -4150,7 +4097,7 @@ "1e-06": "A_receptor_only", "1e-05": "A_receptor_only" }, - "text": "10 nM: A_receptor_only (-0.046); 100 nM: A_receptor_only (-0.348); 1 \u00b5M: A_receptor_only (-0.431); 10 \u00b5M: A_receptor_only (-0.506)", + "text": "10 nM: A_receptor_only (0.095); 100 nM: A_receptor_only (-0.519); 1 \u00b5M: A_receptor_only (-0.564); 10 \u00b5M: A_receptor_only (-0.579)", "unit": null, "note": null, "step": "ablation" @@ -4183,38 +4130,38 @@ }, "abl_concentration_statements": { "value": [ - "Concentration dependence of the ladder: the weakest level is A_receptor_only at every concentration tested (1e-08 M: A_receptor_only (-0.046), 1e-07 M: A_receptor_only (-0.348), 1e-06 M: A_receptor_only (-0.431), 1e-05 M: A_receptor_only (-0.506)).", + "Concentration dependence of the ladder: the weakest level is A_receptor_only at every concentration tested (1e-08 M: A_receptor_only (0.095), 1e-07 M: A_receptor_only (-0.519), 1e-06 M: A_receptor_only (-0.564), 1e-05 M: A_receptor_only (-0.579)).", "With the historical depression-only level-C rule the weakest level is 1e-08 M: C_topology_only, 1e-07 M: A_receptor_only, 1e-06 M: A_receptor_only, 1e-05 M: A_receptor_only: that rule is the worst of the four only at the lowest concentration, and at the three higher ones the receptor-only level is worse. Any sentence of the form 'the topology-only level is the worst of the four' is therefore true of one column of this table and false of the other three." ], - "text": "Concentration dependence of the ladder: the weakest level is A_receptor_only at every concentration tested (1e-08 M: A_receptor_only (-0.046), 1e-07 M: A_receptor_only (-0.348), 1e-06 M: A_receptor_only (-0.431), 1e-05 M: A_receptor_only (-0.506)). With the historical depression-only level-C rule the weakest level is 1e-08 M: C_topology_only, 1e-07 M: A_receptor_only, 1e-06 M: A_receptor_only, 1e-05 M: A_receptor_only: that rule is the worst of the four only at the lowest concentration, and at the three higher ones the receptor-only level is worse. Any sentence of the form 'the topology-only level is the worst of the four' is therefore true of one column of this table and false of the other three.", + "text": "Concentration dependence of the ladder: the weakest level is A_receptor_only at every concentration tested (1e-08 M: A_receptor_only (0.095), 1e-07 M: A_receptor_only (-0.519), 1e-06 M: A_receptor_only (-0.564), 1e-05 M: A_receptor_only (-0.579)). With the historical depression-only level-C rule the weakest level is 1e-08 M: C_topology_only, 1e-07 M: A_receptor_only, 1e-06 M: A_receptor_only, 1e-05 M: A_receptor_only: that rule is the worst of the four only at the lowest concentration, and at the three higher ones the receptor-only level is worse. Any sentence of the form 'the topology-only level is the worst of the four' is therefore true of one column of this table and false of the other three.", "unit": null, "note": null, "step": "ablation" }, "abl_nicotinic_n": { - "value": 9, - "text": "9", + "value": 7, + "text": "7", "unit": null, "note": null, "step": "ablation" }, "abl_rho_composition_nicotinic": { - "value": 0.9534625892455924, - "text": "0.953", + "value": 0.8846517369293828, + "text": "0.885", "unit": null, "note": null, "step": "ablation" }, "abl_rho_receptor_nicotinic": { - "value": -0.7866339502806552, - "text": "-0.787", + "value": -0.8747670497362986, + "text": "-0.875", "unit": null, "note": null, "step": "ablation" }, "abl_ref_observed": { - "value": 0.9878103669773378, - "text": "0.988", + "value": 0.9214816983710898, + "text": "0.921", "unit": null, "note": null, "step": "ablation" @@ -4255,15 +4202,15 @@ "step": "ablation" }, "abl_ref_observed_percentile": { - "value": 100.0, - "text": "100", + "value": 80.0, + "text": "80.0", "unit": null, "note": null, "step": "ablation" }, "abl_ref_shuffled": { - "value": 0.9878103669773378, - "text": "0.988", + "value": 0.9214816983710898, + "text": "0.921", "unit": null, "note": null, "step": "ablation" @@ -4276,36 +4223,36 @@ "step": "ablation" }, "abl_ref_no_floor": { - "value": 0.9892086330935252, - "text": "0.989", + "value": 0.8263565891472867, + "text": "0.826", "unit": null, "note": null, "step": "ablation" }, "abl_ref_no_floor_n": { - "value": 15, - "text": "15", + "value": 16, + "text": "16", "unit": null, "note": null, "step": "ablation" }, "abl_ref_statements": { "value": [ - "Composition-versus-full Spearman rho = 0.988 over 21 library compounds at 1.0e-06 M.", - "Dropping the 6 compound(s) whose gain vector sits on the 0.05 floor leaves rho = 0.989: the correlation is not a tie artefact.", - "Reference 1 (arbitrary gain vectors, all gains moved at once): 50 draws of 21 vectors give rho median 0.808, 5th-95th percentile [0.615, 0.907]; the observed 0.988 is at the 100th percentile. Even compounds that do not exist, ordered by nothing but a shared random gain vector, reproduce most of the ordering.", - "Reference 2 (matched: one gain moved per pseudo-compound, the shape the shipped mechanism rules produce): rho median 0.847, 5th-95th percentile [0.676, 0.953] over 50 draws. The observed 0.988 is at the 100th percentile - above the reference range, but the reference median is already 0.847. A rank correlation of this size between levels B and D is therefore close to an algebraic identity: both are functions of the same gain vector, one through a fixed census and one through a row-normalised matrix whose cells average their presynaptic gains.", - "WHAT WOULD HAVE BEEN EVIDENCE AGAINST THE CLAIM that composition carries the ordering: an observed rho below the matched reference's 5th percentile (0.676), i.e. the census ordering the REAL library worse than it orders pseudo-compounds with no pharmacology in them. Evidence that the agreement was a substantive finding about this connectome would need the matched reference to sit well below the observed value - a median near zero, so that reproducing the ordering was a priori unlikely. Here the matched reference median is 0.847, so the quoted 0.99 should be read as the structural floor plus a small excess, not as a measurement of what the connectome contributes.", - "Shuffling the compound-to-gain assignment leaves rho = 0.988 against the observed 0.988 (difference +0). Both levels are functions of the gain vector alone, so a permutation of the labels moves the (B, D) pairs as a block and cannot change the rank correlation. The comparison therefore carries no compound-level information: it is a statement about the map from gains to readouts, not about which compound is which." + "Composition-versus-full Spearman rho = 0.921 over 21 library compounds at 1.0e-06 M.", + "Dropping the 5 compound(s) whose gain vector sits on the 0.05 floor leaves rho = 0.826: the correlation is not a tie artefact.", + "Reference 1 (arbitrary gain vectors, all gains moved at once): 50 draws of 21 vectors give rho median 0.808, 5th-95th percentile [0.615, 0.907]; the observed 0.921 is at the 96th percentile. Even compounds that do not exist, ordered by nothing but a shared random gain vector, reproduce most of the ordering.", + "Reference 2 (matched: one gain moved per pseudo-compound, the shape the shipped mechanism rules produce): rho median 0.847, 5th-95th percentile [0.676, 0.953] over 50 draws. The observed 0.921 is at the 80th percentile - UNSURPRISING: it lies inside the reference's 5th-95th percentile. A high correlation between levels B and D can arise because both are functions of the same gain vector, one through a fixed census and one through a row-normalised matrix whose cells average their presynaptic gains.", + "The observed value should be read against the matched reference interval [0.676, 0.953], whose median is 0.847. A large B-versus-D correlation alone does not measure a compound-specific contribution of the connectome.", + "Shuffling the compound-to-gain assignment leaves rho = 0.921 against the observed 0.921 (difference +0). Both levels are functions of the gain vector alone, so a permutation of the labels moves the (B, D) pairs as a block and cannot change the rank correlation. This label shuffle supplies no independent compound-level evidence; it tests the shared gain-to-readout map." ], - "text": "Composition-versus-full Spearman rho = 0.988 over 21 library compounds at 1.0e-06 M. Dropping the 6 compound(s) whose gain vector sits on the 0.05 floor leaves rho = 0.989: the correlation is not a tie artefact. Reference 1 (arbitrary gain vectors, all gains moved at once): 50 draws of 21 vectors give rho median 0.808, 5th-95th percentile [0.615, 0.907]; the observed 0.988 is at the 100th percentile. Even compounds that do not exist, ordered by nothing but a shared random gain vector, reproduce most of the ordering. Reference 2 (matched: one gain moved per pseudo-compound, the shape the shipped mechanism rules produce): rho median 0.847, 5th-95th percentile [0.676, 0.953] over 50 draws. The observed 0.988 is at the 100th percentile - above the reference range, but the reference median is already 0.847. A rank correlation of this size between levels B and D is therefore close to an algebraic identity: both are functions of the same gain vector, one through a fixed census and one through a row-normalised matrix whose cells average their presynaptic gains. WHAT WOULD HAVE BEEN EVIDENCE AGAINST THE CLAIM that composition carries the ordering: an observed rho below the matched reference's 5th percentile (0.676), i.e. the census ordering the REAL library worse than it orders pseudo-compounds with no pharmacology in them. Evidence that the agreement was a substantive finding about this connectome would need the matched reference to sit well below the observed value - a median near zero, so that reproducing the ordering was a priori unlikely. Here the matched reference median is 0.847, so the quoted 0.99 should be read as the structural floor plus a small excess, not as a measurement of what the connectome contributes. Shuffling the compound-to-gain assignment leaves rho = 0.988 against the observed 0.988 (difference +0). Both levels are functions of the gain vector alone, so a permutation of the labels moves the (B, D) pairs as a block and cannot change the rank correlation. The comparison therefore carries no compound-level information: it is a statement about the map from gains to readouts, not about which compound is which.", + "text": "Composition-versus-full Spearman rho = 0.921 over 21 library compounds at 1.0e-06 M. Dropping the 5 compound(s) whose gain vector sits on the 0.05 floor leaves rho = 0.826: the correlation is not a tie artefact. Reference 1 (arbitrary gain vectors, all gains moved at once): 50 draws of 21 vectors give rho median 0.808, 5th-95th percentile [0.615, 0.907]; the observed 0.921 is at the 96th percentile. Even compounds that do not exist, ordered by nothing but a shared random gain vector, reproduce most of the ordering. Reference 2 (matched: one gain moved per pseudo-compound, the shape the shipped mechanism rules produce): rho median 0.847, 5th-95th percentile [0.676, 0.953] over 50 draws. The observed 0.921 is at the 80th percentile - UNSURPRISING: it lies inside the reference's 5th-95th percentile. A high correlation between levels B and D can arise because both are functions of the same gain vector, one through a fixed census and one through a row-normalised matrix whose cells average their presynaptic gains. The observed value should be read against the matched reference interval [0.676, 0.953], whose median is 0.847. A large B-versus-D correlation alone does not measure a compound-specific contribution of the connectome. Shuffling the compound-to-gain assignment leaves rho = 0.921 against the observed 0.921 (difference +0). Both levels are functions of the gain vector alone, so a permutation of the labels moves the (B, D) pairs as a block and cannot change the rank correlation. This label shuffle supplies no independent compound-level evidence; it tests the shared gain-to-readout map.", "unit": null, "note": null, "step": "ablation" }, "abl_rho_composition_row_abs": { - "value": 0.9878103669773378, - "text": "0.988", + "value": 0.9214816983710898, + "text": "0.921", "unit": null, "note": null, "step": "ablation" @@ -4318,8 +4265,8 @@ "step": "ablation" }, "abl_rho_composition_none": { - "value": 0.9056122378744635, - "text": "0.906", + "value": 0.9348816268044955, + "text": "0.935", "unit": null, "note": null, "step": "ablation" @@ -4332,8 +4279,8 @@ "step": "ablation" }, "abl_rho_composition_degree": { - "value": 0.6662469964157954, - "text": "0.666", + "value": 0.5510525399252045, + "text": "0.551", "unit": null, "note": null, "step": "ablation" @@ -4346,22 +4293,22 @@ "step": "ablation" }, "abl_rho_composition_row_abs_1e8": { - "value": 0.9061889250814332, - "text": "0.906", + "value": 0.8763297872340425, + "text": "0.876", "unit": null, "note": null, "step": "ablation" }, "abl_rho_composition_none_1e8": { - "value": 0.9214681776227494, - "text": "0.921", + "value": 0.9091349346404287, + "text": "0.909", "unit": null, "note": null, "step": "ablation" }, "abl_rho_composition_degree_1e8": { - "value": -0.37198697068403913, - "text": "-0.372", + "value": -0.5832082533976854, + "text": "-0.583", "unit": null, "note": null, "step": "ablation" @@ -4378,30 +4325,30 @@ }, "abl_normalisation_statements": { "value": [ - "Composition-versus-full Spearman rho at 1.0e-06 M: row_abs 0.988, none 0.906, degree 0.666.", - "The composition-dominated verdict survives only under ['row_abs', 'none'] (threshold rho >= 0.9); the shipped engine is 'row_abs', and the row normalisation is an undocumented modelling choice, not a measurement." + "Composition-versus-full Spearman rho at 1.0e-06 M: row_abs 0.921, none 0.935, degree 0.551.", + "The composition-dominated verdict survives only under ['row_abs', 'none'] (threshold rho >= 0.9); the shipped engine is 'row_abs', and the row normalisation is a documented modelling choice, not a measurement." ], - "text": "Composition-versus-full Spearman rho at 1.0e-06 M: row_abs 0.988, none 0.906, degree 0.666. The composition-dominated verdict survives only under ['row_abs', 'none'] (threshold rho >= 0.9); the shipped engine is 'row_abs', and the row normalisation is an undocumented modelling choice, not a measurement.", + "text": "Composition-versus-full Spearman rho at 1.0e-06 M: row_abs 0.921, none 0.935, degree 0.551. The composition-dominated verdict survives only under ['row_abs', 'none'] (threshold rho >= 0.9); the shipped engine is 'row_abs', and the row normalisation is a documented modelling choice, not a measurement.", "unit": null, "note": null, "step": "ablation" }, "abl_glutamate_max_delta_rho": { - "value": 0.05732899022801308, - "text": "0.057", + "value": 0.08643617021276595, + "text": "0.086", "unit": null, "note": null, "step": "ablation" }, "abl_glutamate_statements": { "value": [ - "At 1.0e-08 M the composition-versus-full rho is 0.906 with glutamate scored partly excitatory (shipped) and 0.964 with glutamate signed as the rate engine signs it (+0.057).", - "At 1.0e-07 M the composition-versus-full rho is 0.989 with glutamate scored partly excitatory (shipped) and 0.987 with glutamate signed as the rate engine signs it (-0.001).", - "At 1.0e-06 M the composition-versus-full rho is 0.988 with glutamate scored partly excitatory (shipped) and 0.984 with glutamate signed as the rate engine signs it (-0.004).", - "At 1.0e-05 M the composition-versus-full rho is 0.987 with glutamate scored partly excitatory (shipped) and 0.979 with glutamate signed as the rate engine signs it (-0.008).", - "The two levels of the ladder disagree about the sign of glutamate; reconciling them moves the number the paper's conclusion rests on by at most 0.057 in rank correlation. The disagreement is a real defect of the ladder and is documented as one, but it is not what produces the composition-versus-full agreement." + "At 1.0e-08 M the composition-versus-full rho is 0.876 with glutamate scored partly excitatory (shipped) and 0.963 with glutamate signed as the rate engine signs it (+0.086).", + "At 1.0e-07 M the composition-versus-full rho is 0.918 with glutamate scored partly excitatory (shipped) and 0.992 with glutamate signed as the rate engine signs it (+0.073).", + "At 1.0e-06 M the composition-versus-full rho is 0.921 with glutamate scored partly excitatory (shipped) and 0.991 with glutamate signed as the rate engine signs it (+0.070).", + "At 1.0e-05 M the composition-versus-full rho is 0.954 with glutamate scored partly excitatory (shipped) and 0.992 with glutamate signed as the rate engine signs it (+0.038).", + "The two levels of the ladder disagree about the sign of glutamate; reconciling them moves the number the paper's conclusion rests on by at most 0.086 in rank correlation. The disagreement is a real defect of the ladder and is documented as one, but it is not what produces the composition-versus-full agreement." ], - "text": "At 1.0e-08 M the composition-versus-full rho is 0.906 with glutamate scored partly excitatory (shipped) and 0.964 with glutamate signed as the rate engine signs it (+0.057). At 1.0e-07 M the composition-versus-full rho is 0.989 with glutamate scored partly excitatory (shipped) and 0.987 with glutamate signed as the rate engine signs it (-0.001). At 1.0e-06 M the composition-versus-full rho is 0.988 with glutamate scored partly excitatory (shipped) and 0.984 with glutamate signed as the rate engine signs it (-0.004). At 1.0e-05 M the composition-versus-full rho is 0.987 with glutamate scored partly excitatory (shipped) and 0.979 with glutamate signed as the rate engine signs it (-0.008). The two levels of the ladder disagree about the sign of glutamate; reconciling them moves the number the paper's conclusion rests on by at most 0.057 in rank correlation. The disagreement is a real defect of the ladder and is documented as one, but it is not what produces the composition-versus-full agreement.", + "text": "At 1.0e-08 M the composition-versus-full rho is 0.876 with glutamate scored partly excitatory (shipped) and 0.963 with glutamate signed as the rate engine signs it (+0.086). At 1.0e-07 M the composition-versus-full rho is 0.918 with glutamate scored partly excitatory (shipped) and 0.992 with glutamate signed as the rate engine signs it (+0.073). At 1.0e-06 M the composition-versus-full rho is 0.921 with glutamate scored partly excitatory (shipped) and 0.991 with glutamate signed as the rate engine signs it (+0.070). At 1.0e-05 M the composition-versus-full rho is 0.954 with glutamate scored partly excitatory (shipped) and 0.992 with glutamate signed as the rate engine signs it (+0.038). The two levels of the ladder disagree about the sign of glutamate; reconciling them moves the number the paper's conclusion rests on by at most 0.086 in rank correlation. The disagreement is a real defect of the ladder and is documented as one, but it is not what produces the composition-versus-full agreement.", "unit": null, "note": null, "step": "ablation" @@ -4421,22 +4368,22 @@ "step": "landscape" }, "landscape_scored": { - "value": 20, - "text": "20", + "value": 16, + "text": "16", "unit": null, "note": null, "step": "landscape" }, "landscape_compounds_scored": { - "value": 10, - "text": "10", + "value": 8, + "text": "8", "unit": null, "note": null, "step": "landscape" }, "landscape_amplify_rows": { - "value": 8, - "text": "8", + "value": 4, + "text": "4", "unit": null, "note": null, "step": "landscape" @@ -4450,12 +4397,10 @@ }, "landscape_amplify_compounds": { "value": [ - "chlorpyrifos_oxon", "ddt", - "deltamethrin", - "permethrin" + "deltamethrin" ], - "text": "chlorpyrifos_oxon, ddt, deltamethrin, permethrin", + "text": "ddt, deltamethrin", "unit": null, "note": null, "step": "landscape" @@ -4475,8 +4420,8 @@ "step": "landscape" }, "landscape_amplify_mean_gap": { - "value": 0.2864971832082256, - "text": "0.286", + "value": 0.19141337118506263, + "text": "0.191", "unit": null, "note": null, "step": "landscape" @@ -4489,8 +4434,8 @@ "step": "landscape" }, "landscape_gap_split": { - "value": 1.1563562211532965, - "text": "1.156", + "value": 1.0612724091301335, + "text": "1.061", "unit": null, "note": null, "step": "landscape" @@ -4502,12 +4447,11 @@ "fipronil", "gaba", "ivermectin", + "permethrin", "picrotoxin", - "spinosad", - "sulfoxaflor", "thiamethoxam" ], - "text": "chlordimeform, dieldrin, fipronil, gaba, ivermectin, picrotoxin, spinosad, sulfoxaflor, thiamethoxam", + "text": "chlordimeform, dieldrin, fipronil, gaba, ivermectin, permethrin, picrotoxin, thiamethoxam", "unit": null, "note": null, "step": "landscape" @@ -4541,7 +4485,7 @@ "step": "dose" }, "ic50_hi_M": { - "value": 7.15055024256253e-08, + "value": 7.15055024856943e-08, "text": "71.5 nM", "unit": null, "note": null, @@ -4564,14 +4508,14 @@ "ic50_top_hz": { "value": 6.689915702519358, "text": "6.690", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dose" }, "ic50_bottom_hz": { "value": 0.45604816528092096, "text": "0.456", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dose" }, @@ -4599,35 +4543,35 @@ "tornado_span_gain_coef": { "value": 2.239375785860553, "text": "2.239", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dose" }, "tornado_span_drive_hz": { "value": 0.6876931796887389, "text": "0.688", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dose" }, "tornado_span_weight_threshold": { "value": 0.19864315539146077, "text": "0.199", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dose" }, "tornado_span_ec50": { "value": 0.0, "text": "0.000", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dose" }, "tornado_span_hill_n": { "value": 0.0, "text": "0.000", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "dose" }, @@ -5193,8 +5137,8 @@ "step": "predictions" }, "pred_H7_effect": { - "value": 0.3397826600493299, - "text": "0.340", + "value": 0.40019619368267084, + "text": "0.400", "unit": null, "note": null, "step": "predictions" @@ -5290,104 +5234,6 @@ "note": null, "step": "mixtures" }, - "expr_coverage_overall": { - "value": 0.2053252443545669, - "text": "0.205", - "unit": null, - "note": null, - "step": "expression" - }, - "expr_cov_named_insect_nAChR": { - "value": 0.458259325044405, - "text": "0.458", - "unit": null, - "note": null, - "step": "expression" - }, - "expr_cov_named_insect_RDL": { - "value": 0.6563055062166963, - "text": "0.656", - "unit": null, - "note": null, - "step": "expression" - }, - "expr_cov_named_insect_GluCl": { - "value": 0.0719360568383659, - "text": "0.072", - "unit": null, - "note": null, - "step": "expression" - }, - "expr_cov_named_insect_OctR": { - "value": 0.0, - "text": "0.000", - "unit": null, - "note": null, - "step": "expression" - }, - "expr_cov_named_insect_AChE": { - "value": 0.0, - "text": "0.000", - "unit": null, - "note": null, - "step": "expression" - }, - "expr_rows": { - "value": 50, - "text": "50", - "unit": null, - "note": null, - "step": "expression" - }, - "expr_motor_gap": { - "value": true, - "text": "yes", - "unit": null, - "note": null, - "step": "expression" - }, - "expr_motor_gap_note": { - "value": "motor-neuron receptor expression is a confirmed gap", - "text": "motor-neuron receptor expression is a confirmed gap", - "unit": null, - "note": null, - "step": "expression" - }, - "expr_mn9_uniform_hz": { - "value": 39.40567922340328, - "text": "39.4", - "unit": "Hz", - "note": null, - "step": "expression" - }, - "expr_mn9_weighted_hz": { - "value": 36.78411732719066, - "text": "36.8", - "unit": "Hz", - "note": null, - "step": "expression" - }, - "expr_mn9_delta_hz": { - "value": -2.6215618962126257, - "text": "-2.622", - "unit": "Hz", - "note": null, - "step": "expression" - }, - "expr_mean_uniform_hz": { - "value": 0.45846211979249263, - "text": "0.458", - "unit": "Hz", - "note": null, - "step": "expression" - }, - "expr_mean_weighted_hz": { - "value": 0.8585976869666511, - "text": "0.859", - "unit": "Hz", - "note": null, - "step": "expression" - }, "stab_family_size": { "value": 25, "text": "25", @@ -5610,15 +5456,15 @@ "step": "stability" }, "stab_C6_retained": { - "value": 19, - "text": "19", + "value": 18, + "text": "18", "unit": null, "note": null, "step": "stability" }, "stab_C6_lost": { - "value": 6, - "text": "6", + "value": 7, + "text": "7", "unit": null, "note": null, "step": "stability" @@ -5631,8 +5477,8 @@ "step": "stability" }, "stab_C6_fraction": { - "value": 0.76, - "text": "0.760", + "value": 0.72, + "text": "0.720", "unit": null, "note": null, "step": "stability" @@ -5644,9 +5490,10 @@ "strong_biphasic@0.50x", "strong_biphasic@0.75x", "strong_biphasic@1.00x", - "weak_biphasic@0.50x" + "weak_biphasic@0.50x", + "weak_biphasic@0.75x" ], - "text": "flylab_biphasic@0.50x, linear@0.50x, strong_biphasic@0.50x, strong_biphasic@0.75x, strong_biphasic@1.00x, weak_biphasic@0.50x", + "text": "flylab_biphasic@0.50x, linear@0.50x, strong_biphasic@0.50x, strong_biphasic@0.75x, strong_biphasic@1.00x, weak_biphasic@0.50x, weak_biphasic@0.75x", "unit": null, "note": null, "step": "stability" @@ -5985,14 +5832,14 @@ "stab_default_treated_hz": { "value": 0.45846211979249263, "text": "0.458", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "stability" }, "stab_default_vehicle_hz": { "value": 6.631244285926711, "text": "6.631", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "stability" }, @@ -6004,16 +5851,16 @@ "step": "stability" }, "stab_monotone_treated_hz": { - "value": 10.103618005735553, + "value": 10.103618005735555, "text": "10.1", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "stability" }, "stab_monotone_vehicle_hz": { "value": 6.631244285926711, "text": "6.631", - "unit": "Hz", + "unit": "rate-model units", "note": null, "step": "stability" }, @@ -6025,7 +5872,7 @@ "step": "stability" }, "stab_monotone_pct_change": { - "value": 52.36383354445493, + "value": 52.36383354445496, "text": "+52%", "unit": null, "note": null, @@ -6039,7 +5886,7 @@ "step": "stability" }, "stab_mean_gap_nicotinic_monotone": { - "value": -0.8929449323721664, + "value": -0.8929449323721661, "text": "-0.893", "unit": null, "note": null, @@ -6053,7 +5900,7 @@ "step": "stability" }, "stab_mean_gap_nicotinic_monotone_min": { - "value": -1.7336718357793948, + "value": -1.7336718357793934, "text": "-1.734", "unit": null, "note": null, @@ -6119,25 +5966,25 @@ "thr_t15_scored": { "value": [ "acetamiprid", - "chlorpyrifos_oxon", "clothianidin", "ddt", "deltamethrin", "imidacloprid", "nicotine", - "nitenpyram", - "permethrin" + "nitenpyram" ], - "text": "acetamiprid, chlorpyrifos_oxon, clothianidin, ddt, deltamethrin, imidacloprid, nicotine, nitenpyram, permethrin", + "text": "acetamiprid, clothianidin, ddt, deltamethrin, imidacloprid, nicotine, nitenpyram", "unit": null, "note": null, "step": "stability" }, "thr_t15_unscored": { "value": [ + "chlorpyrifos_oxon", + "permethrin", "spinosad" ], - "text": "spinosad", + "text": "chlorpyrifos_oxon, permethrin, spinosad", "unit": null, "note": null, "step": "stability" @@ -6216,8 +6063,8 @@ "step": "stability" }, "thr_nav_ache_matching": { - "value": 8, - "text": "8", + "value": 7, + "text": "7", "unit": null, "note": null, "step": "stability" @@ -6238,9 +6085,10 @@ }, "thr_nav_ache_failing": { "value": [ + "circuit_frac=0.40,vert_limit=0.10", "circuit_frac=0.50,vert_limit=0.10" ], - "text": "circuit_frac=0.50,vert_limit=0.10", + "text": "circuit_frac=0.40,vert_limit=0.10, circuit_frac=0.50,vert_limit=0.10", "unit": null, "note": null, "step": "stability" @@ -6253,1210 +6101,1309 @@ "step": "stability" }, "thr_nicotinic_gap_max": { - "value": -0.362108493778807, + "value": -0.36210849377880716, "text": "-0.362", "unit": null, "note": null, "step": "stability" }, "thr_nav_ache_gap_min": { - "value": -0.0752461538175645, - "text": "-0.075", + "value": -0.15388276149595248, + "text": "-0.154", "unit": null, "note": null, "step": "stability" }, "thr_nav_ache_gap_max": { - "value": 0.8131579176045398, - "text": "0.813", + "value": 0.7524306974807679, + "text": "0.752", "unit": null, "note": null, "step": "stability" }, - "unc_n_base": { - "value": 1024, - "text": "1024", + "scale_ladder_sizes": { + "value": [ + 1000, + 5000, + 10000, + 25000, + 50000 + ], + "text": "1k, 5k, 10k, 25k, 50k", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_n_factors": { - "value": 9, - "text": "9", + "scale_rungs_on_disk": { + "value": [ + "scale_1k" + ], + "text": "scale_1k", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_evaluations": { - "value": 11264, - "text": "11\u202f264", + "scale_rungs_missing": { + "value": [ + "scale_5k", + "scale_10k", + "scale_25k", + "scale_50k" + ], + "text": "scale_5k, scale_10k, scale_25k, scale_50k", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_estimator": { - "value": "jansen1999", - "text": "jansen1999", + "scale_n_rungs_on_disk": { + "value": 1, + "text": "1", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_output_mean": { - "value": 1.4804027560720454, - "text": "1.480", - "unit": "Hz", + "scale_1k_name": { + "value": "scale_1k", + "text": "scale_1k", + "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_variance": { - "value": 3.149230029344893, - "text": "3.149", - "unit": "Hz^2", + "scale_1k_nodes": { + "value": 1000, + "text": "1000", + "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_interaction_share": { - "value": 0.3442830879951747, - "text": "0.344", + "scale_1k_edges": { + "value": 22857, + "text": "22\u202f857", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_interaction_share_clipped": { - "value": 0.27214054387701025, - "text": "0.272", + "scale_1k_mean_degree": { + "value": 22.857, + "text": "22.9", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_sum_first_order": { - "value": 0.6557169120048253, - "text": "0.656", + "scale_1k_share_onto_seeds": { + "value": 0.06501290633066456, + "text": "7%", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_sum_first_order_clipped": { - "value": 0.7278594561229897, - "text": "0.728", + "scale_1k_in_star": { + "value": false, + "text": "no", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_S_gain_transform": { - "value": 0.4096497698742678, - "text": "0.410", + "scale_frontier_n_paper": { + "value": 1000, + "text": "1000", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_ST_gain_transform": { - "value": 0.6264806479027828, - "text": "0.626", + "scale_largest_feasible_profile": { + "value": "scale_50k", + "text": "scale_50k", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_CI_gain_transform": { - "value": [ - 0.31256086807181116, - 0.5001624331362007 - ], - "text": "(+0.313, +0.500)", + "scale_largest_feasible_lif": { + "value": "scale_50k", + "text": "scale_50k", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_S_weight_threshold": { - "value": 0.29357557224259245, - "text": "0.294", + "scale_top_rung": { + "value": "50k", + "text": "50k", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_ST_weight_threshold": { - "value": 0.5424303725151057, - "text": "0.542", + "scale_top_profile_hours": { + "value": 21.13708841666667, + "text": "21.1", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_CI_weight_threshold": { - "value": [ - 0.1987612666853245, - 0.39130919604916053 - ], - "text": "(+0.199, +0.391)", + "scale_top_landscape_days": { + "value": 7.397980945833334, + "text": "7", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_S_drive": { - "value": 0.018555443445925905, - "text": "0.019", + "scale_whole_cns_profile_hours": { + "value": 59.427331277777775, + "text": "59", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_ST_drive": { - "value": 0.14121674202113985, - "text": "0.141", + "scale_study_source": { + "value": "papers/scale_study.json", + "text": "papers/scale_study.json", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_CI_drive": { - "value": [ - -0.0716609947161836, - 0.11220171357678527 - ], - "text": "(-0.072, +0.112)", + "scale_study_status": { + "value": "reported", + "text": "reported", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_S_gain_coef": { - "value": -0.04565580230843003, - "text": "-0.046", + "scale_study_cuts": { + "value": [ + "scale_1k", + "named", + "taste_motor", + "scale_5k", + "scale_10k", + "scale_25k", + "scale_50k" + ], + "text": "scale_1k, named, taste_motor, scale_5k, scale_10k, scale_25k, scale_50k", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_ST_gain_coef": { - "value": 0.023822223893828185, - "text": "0.024", + "scale_study_n_cuts": { + "value": 7, + "text": "7", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_CI_gain_coef": { + "scale_study_compounds": { "value": [ - -0.10739018244206512, - 0.02765398020264121 + "imidacloprid", + "fipronil" ], - "text": "(-0.107, +0.028)", + "text": "imidacloprid, fipronil", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_S_transmitter": { - "value": 0.006078670560203601, - "text": "0.006", + "scale_study_settled": { + "value": true, + "text": "yes", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_ST_transmitter": { - "value": 0.023642782102253644, - "text": "0.024", + "scale_study_settled_clean_ladder": { + "value": true, + "text": "yes", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_CI_transmitter": { + "scale_study_statement": { "value": [ - -0.05210207681142646, - 0.06965580923052817 + "imidacloprid: the verdict is 'topology-dependent' on the largest 5 of 7 cuts tested, first reached at 'taste_motor'.", + "fipronil: the verdict is 'topology-dependent' on the largest 7 of 7 cuts tested, first reached at 'scale_1k'." ], - "text": "(-0.052, +0.070)", + "text": "imidacloprid: the verdict is 'topology-dependent' on the largest 5 of 7 cuts tested, first reached at 'taste_motor'. fipronil: the verdict is 'topology-dependent' on the largest 7 of 7 cuts tested, first reached at 'scale_1k'.", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_S_hill_n": { - "value": -0.007946568350353221, - "text": "-0.008", + "scale_study_sequences": { + "value": [ + "imidacloprid: scale_1k (topology-dependent) -> scale_5k (topology-dependent) -> scale_10k (topology-dependent) -> scale_25k (topology-dependent) -> scale_50k (topology-dependent)", + "fipronil: scale_1k (topology-dependent) -> scale_5k (topology-dependent) -> scale_10k (topology-dependent) -> scale_25k (topology-dependent) -> scale_50k (topology-dependent)" + ], + "text": "imidacloprid: scale_1k (topology-dependent) -> scale_5k (topology-dependent) -> scale_10k (topology-dependent) -> scale_25k (topology-dependent) -> scale_50k (topology-dependent); fipronil: scale_1k (topology-dependent) -> scale_5k (topology-dependent) -> scale_10k (topology-dependent) -> scale_25k (topology-dependent) -> scale_50k (topology-dependent)", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_ST_hill_n": { - "value": 0.0007525711738826337, - "text": "7.53e-04", + "scale_study_n_range": { + "value": [ + 20, + 1000 + ], + "text": "20-1000", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_CI_hill_n": { + "scale_study_resolution_range": { "value": [ - -0.06210277198208717, - 0.04046709008223071 + 0.000999000999000999, + 0.047619047619047616 ], - "text": "(-0.062, +0.040)", + "text": "0.0010-0.0476", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_S_potency": { - "value": -0.005989859757710479, - "text": "-0.006", + "scale_study_composition_dominated_cuts": { + "value": [ + "named" + ], + "text": "named", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_ST_potency": { - "value": 0.0006784194893530697, - "text": "6.78e-04", + "scale_study_n_composition_dominated": { + "value": 1, + "text": "1", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_CI_potency": { + "scale_study_every_null_cuts": { "value": [ - -0.06115133451219928, - 0.045925733350771346 + "scale_5k", + "scale_10k" ], - "text": "(-0.061, +0.046)", + "text": "scale_5k, scale_10k", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_S_expression": { - "value": -0.005809124285986339, - "text": "-0.006", + "scale_study_coarse_cuts": { + "value": [ + "scale_10k", + "scale_25k", + "scale_50k" + ], + "text": "scale_10k, scale_25k, scale_50k", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_ST_expression": { - "value": 0.00012806053442048645, - "text": "1.28e-04", + "scale_study_coarse_note": { + "value": { + "cuts": [ + "scale_10k", + "scale_25k", + "scale_50k" + ], + "alpha": 0.05 + }, + "text": "at scale_10k, scale_25k, scale_50k the permutation budget falls to n = 20, where the smallest attainable probability is 0.048 against alpha = 0.05: a rejection there is the smallest the test can express", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_CI_expression": { + "scale_study_imidacloprid_sequence": { "value": [ - -0.058783087902346, - 0.044256230041565946 + { + "cut": "scale_1k", + "n_nodes": 1000, + "class": "topology-dependent" + }, + { + "cut": "named", + "n_nodes": 1126, + "class": "composition-dominated" + }, + { + "cut": "taste_motor", + "n_nodes": 1841, + "class": "topology-dependent" + }, + { + "cut": "scale_5k", + "n_nodes": 5000, + "class": "topology-dependent" + }, + { + "cut": "scale_10k", + "n_nodes": 10000, + "class": "topology-dependent" + }, + { + "cut": "scale_25k", + "n_nodes": 25000, + "class": "topology-dependent" + }, + { + "cut": "scale_50k", + "n_nodes": 50000, + "class": "topology-dependent" + } ], - "text": "(-0.059, +0.044)", + "text": "scale_1k (1000 nodes, mean degree 22.9): topology-dependent; named (1126 nodes, mean degree 1.2): composition-dominated; taste_motor (1841 nodes, mean degree 10.4): topology-dependent; scale_5k (5000 nodes, mean degree 56.0): topology-dependent; scale_10k (10000 nodes, mean degree 62.2): topology-dependent; scale_25k (25000 nodes, mean degree 54.6): topology-dependent; scale_50k (50000 nodes, mean degree 44.3): topology-dependent", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_S_lif_seed": { - "value": -0.006741189415684353, - "text": "-0.007", + "scale_study_imidacloprid_relative_range": { + "value": [ + 0.5198156654684973, + 0.9308633342364611 + ], + "text": "0.52-0.93", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_ST_lif_seed": { - "value": 0.0, - "text": "0.000", + "scale_study_fipronil_sequence": { + "value": [ + { + "cut": "scale_1k", + "n_nodes": 1000, + "class": "topology-dependent" + }, + { + "cut": "named", + "n_nodes": 1126, + "class": "topology-dependent" + }, + { + "cut": "taste_motor", + "n_nodes": 1841, + "class": "topology-dependent" + }, + { + "cut": "scale_5k", + "n_nodes": 5000, + "class": "topology-dependent" + }, + { + "cut": "scale_10k", + "n_nodes": 10000, + "class": "topology-dependent" + }, + { + "cut": "scale_25k", + "n_nodes": 25000, + "class": "topology-dependent" + }, + { + "cut": "scale_50k", + "n_nodes": 50000, + "class": "topology-dependent" + } + ], + "text": "scale_1k (1000 nodes, mean degree 22.9): topology-dependent; named (1126 nodes, mean degree 1.2): topology-dependent; taste_motor (1841 nodes, mean degree 10.4): topology-dependent; scale_5k (5000 nodes, mean degree 56.0): topology-dependent; scale_10k (10000 nodes, mean degree 62.2): topology-dependent; scale_25k (25000 nodes, mean degree 54.6): topology-dependent; scale_50k (50000 nodes, mean degree 44.3): topology-dependent", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_CI_lif_seed": { + "scale_study_fipronil_relative_range": { "value": [ - -0.06185506580751919, - 0.043437093161159034 + 0.1356619639942983, + 0.5749207286979054 ], - "text": "(-0.062, +0.043)", + "text": "0.14-0.57", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_top_factor": { - "value": "gain_transform", - "text": "gain_transform", + "scale_study_density_not_size": { + "value": { + "in_star": "named", + "dense": "scale_1k" + }, + "text": "the inversion does not track the number of cells: `scale_1k` has 1000 nodes against `named`'s 1126 and a mean degree of 22.9 against 1.21, and it is already topology-dependent", "unit": null, "note": null, - "step": "uncertainty" + "step": "scale" }, - "unc_second_factor": { - "value": "weight_threshold", - "text": "weight_threshold", + "flylab_version": { + "value": "0.6.0", + "text": "0.6.0", "unit": null, "note": null, - "step": "uncertainty" + "step": "provenance" }, - "unc_noise_floor": { - "value": 0.04565580230843003, - "text": "0.046", + "git_sha": { + "value": "50f5465b2df028a6cb2b93926cfbd14d7c746c9b", + "text": "50f5465b2df0", "unit": null, "note": null, - "step": "uncertainty" + "step": "provenance" }, - "unc_noise_floor_factor": { - "value": "gain_coef", - "text": "gain_coef", + "library_sha256": { + "value": "359096467b33242174dcc5d787a6b7fc6d37573f8daf4771efd159b64ca95e05", + "text": "359096467b33", "unit": null, "note": null, - "step": "uncertainty" + "step": "provenance" }, - "unc_noise_floor_definition": { - "value": "the largest magnitude among the negative first-order estimates, which is a measured lower bound on the estimator's error because a true first-order index cannot be negative; it is compared against the null factor's own index and the larger is used", - "text": "the largest magnitude among the negative first-order estimates, which is a measured lower bound on the estimator's error because a true first-order index cannot be negative; it is compared against the null factor's own index and the larger is used", + "map_id": { + "value": "male-cns:v1.0", + "text": "male-cns:v1.0", + "unit": null, + "note": null, + "step": "provenance" + }, + "named_nodes": { + "value": 1126, + "text": "1126", + "unit": null, + "note": null, + "step": "provenance" + }, + "named_edges": { + "value": 1360, + "text": "1360", + "unit": null, + "note": null, + "step": "provenance" + }, + "taste_nodes": { + "value": 1841, + "text": "1841", + "unit": null, + "note": null, + "step": "provenance" + }, + "taste_edges": { + "value": 19066, + "text": "19\u202f066", + "unit": null, + "note": null, + "step": "provenance" + }, + "named_min_weight": { + "value": 5, + "text": "5", + "unit": null, + "note": null, + "step": "provenance" + }, + "taste_closure_min_weight": { + "value": 10, + "text": "10", + "unit": null, + "note": null, + "step": "provenance" + }, + "expr_coverage_overall": { + "value": 0.2053252443545669, + "text": "0.205", + "unit": null, + "note": null, + "step": "expression" + }, + "expr_cov_named_insect_nAChR": { + "value": 0.458259325044405, + "text": "0.458", + "unit": null, + "note": null, + "step": "expression" + }, + "expr_cov_named_insect_RDL": { + "value": 0.6563055062166963, + "text": "0.656", + "unit": null, + "note": null, + "step": "expression" + }, + "expr_cov_named_insect_GluCl": { + "value": 0.0719360568383659, + "text": "0.072", + "unit": null, + "note": null, + "step": "expression" + }, + "expr_cov_named_insect_OctR": { + "value": 0.0, + "text": "0.000", + "unit": null, + "note": null, + "step": "expression" + }, + "expr_cov_named_insect_AChE": { + "value": 0.0, + "text": "0.000", + "unit": null, + "note": null, + "step": "expression" + }, + "expr_rows": { + "value": 50, + "text": "50", "unit": null, "note": null, - "step": "uncertainty" + "step": "expression" }, - "unc_null_factor_S": { - "value": -0.006741189415684353, - "text": "-0.007", + "expr_motor_gap": { + "value": true, + "text": "yes", "unit": null, "note": null, - "step": "uncertainty" + "step": "expression" }, - "unc_resolved": { - "value": [ - "gain_transform", - "weight_threshold" - ], - "text": "gain_transform, weight_threshold", + "expr_motor_gap_note": { + "value": "motor-neuron receptor expression is a confirmed gap", + "text": "motor-neuron receptor expression is a confirmed gap", "unit": null, "note": null, - "step": "uncertainty" + "step": "expression" }, - "unc_n_resolved": { - "value": 2, - "text": "2", - "unit": null, + "expr_mn9_uniform_hz": { + "value": 39.40567922340328, + "text": "39.4", + "unit": "rate-model units", "note": null, - "step": "uncertainty" + "step": "expression" }, - "unc_at_noise_floor": { - "value": [ - "drive", - "expression", - "gain_coef", - "hill_n", - "lif_seed", - "potency", - "transmitter" - ], - "text": "drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter", - "unit": null, + "expr_mn9_weighted_hz": { + "value": 36.78411732719066, + "text": "36.8", + "unit": "rate-model units", "note": null, - "step": "uncertainty" + "step": "expression" }, - "unc_null_control_factors": { - "value": [ - "drive", - "expression", - "gain_coef", - "hill_n", - "lif_seed", - "potency", - "transmitter" - ], - "text": "drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter", - "unit": null, + "expr_mn9_delta_hz": { + "value": -2.6215618962126257, + "text": "-2.622", + "unit": "rate-model units", "note": null, - "step": "uncertainty" + "step": "expression" }, - "unc_resolution_statement": { - "value": "first-order confidence intervals exclude zero for gain_transform (S1 +0.410, CI [+0.313, +0.500]), weight_threshold (S1 +0.294, CI [+0.199, +0.391]); every other effect is unresolved at this sample size (drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter). 'Unresolved' means the sample cannot separate the effect from zero; it is not a claim that the effect is zero, and a negative point estimate is estimator noise, not a negative contribution. Indistinguishable from the estimator's own noise (floor +0.046, set by gain_coef; the deliberately null factor lif_seed estimates -0.007): drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter.", - "text": "first-order confidence intervals exclude zero for gain_transform (S1 +0.410, CI [+0.313, +0.500]), weight_threshold (S1 +0.294, CI [+0.199, +0.391]); every other effect is unresolved at this sample size (drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter). 'Unresolved' means the sample cannot separate the effect from zero; it is not a claim that the effect is zero, and a negative point estimate is estimator noise, not a negative contribution. Indistinguishable from the estimator's own noise (floor +0.046, set by gain_coef; the deliberately null factor lif_seed estimates -0.007): drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter.", - "unit": null, + "expr_mean_uniform_hz": { + "value": 0.45846211979249263, + "text": "0.458", + "unit": "rate-model units", "note": null, - "step": "uncertainty" + "step": "expression" }, - "unc_max_delta_first": { - "value": 0.04034953984299405, - "text": "0.040", + "expr_mean_weighted_hz": { + "value": 0.8585976869666511, + "text": "0.859", + "unit": "rate-model units", + "note": null, + "step": "expression" + }, + "unc_n_base": { + "value": 1024, + "text": "1024", "unit": null, "note": null, "step": "uncertainty" }, - "ish_n_base": { - "value": 16384, - "text": "16\u202f384", + "unc_n_factors": { + "value": 9, + "text": "9", "unit": null, "note": null, "step": "uncertainty" }, - "ish_estimated": { - "value": [ - 0.3163, - 0.4495, - 0.0072 - ], - "text": "0.3163, 0.4495, 0.0072", + "unc_evaluations": { + "value": 11264, + "text": "11\u202f264", "unit": null, "note": null, "step": "uncertainty" }, - "ish_analytic": { - "value": [ - 0.3139, - 0.4424, - 0.0 - ], - "text": "0.3139, 0.4424, 0.0000", + "unc_estimator": { + "value": "jansen1999", + "text": "jansen1999", "unit": null, "note": null, "step": "uncertainty" }, - "ish_max_abs_error": { - "value": 0.007165317322344709, - "text": "0.007", - "unit": null, + "unc_output_mean": { + "value": 1.4804027560720454, + "text": "1.480", + "unit": "rate-model units", "note": null, "step": "uncertainty" }, - "voi_var_total": { + "unc_variance": { "value": 3.149230029344893, "text": "3.149", - "unit": "Hz^2", + "unit": "rate-model units^2", "note": null, "step": "uncertainty" }, - "voi_rank1_factor": { - "value": "gain_transform", - "text": "gain_transform", + "unc_interaction_share": { + "value": 0.3442830879951748, + "text": "0.344", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank1_var": { - "value": 1.290081356802269, - "text": "1.290", - "unit": "Hz^2", - "note": null, - "step": "uncertainty" - }, - "voi_rank1_fraction": { - "value": 0.4096497698742678, - "text": "0.410", + "unc_interaction_share_clipped": { + "value": 0.27214054387701025, + "text": "0.272", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank1_experiment": { - "value": "calibration of receptor engagement against synaptic gain", - "text": "calibration of receptor engagement against synaptic gain", + "unc_sum_first_order": { + "value": 0.6557169120048252, + "text": "0.656", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank1_cost": { - "value": "high", - "text": "high", + "unc_sum_first_order_clipped": { + "value": 0.7278594561229897, + "text": "0.728", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank1_state": { - "value": "resolved", - "text": "resolved", + "unc_S_gain_transform": { + "value": 0.4096497698742678, + "text": "0.410", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank2_factor": { - "value": "weight_threshold", - "text": "weight_threshold", + "unc_ST_gain_transform": { + "value": 0.6264806479027828, + "text": "0.626", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank2_var": { - "value": 0.9245370079884832, - "text": "0.925", - "unit": "Hz^2", + "unc_CI_gain_transform": { + "value": [ + 0.31256086807181116, + 0.5001624331362007 + ], + "text": "(+0.313, +0.500)", + "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank2_fraction": { + "unc_S_weight_threshold": { "value": 0.29357557224259245, "text": "0.294", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank2_experiment": { - "value": "no experiment: a reconstruction-confidence analysis", - "text": "no experiment: a reconstruction-confidence analysis", + "unc_ST_weight_threshold": { + "value": 0.5424303725151058, + "text": "0.542", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank2_cost": { - "value": "none (compute only)", - "text": "none (compute only)", + "unc_CI_weight_threshold": { + "value": [ + 0.1987612666853245, + 0.39130919604916053 + ], + "text": "(+0.199, +0.391)", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank2_state": { - "value": "resolved", - "text": "resolved", + "unc_S_drive": { + "value": 0.018555443445925905, + "text": "0.019", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank3_factor": { - "value": "drive", - "text": "drive", + "unc_ST_drive": { + "value": 0.14121674202113985, + "text": "0.141", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank3_var": { - "value": 0.05843535970772074, - "text": "0.058", - "unit": "Hz^2", - "note": null, - "step": "uncertainty" - }, - "voi_rank3_fraction": { - "value": 0.018555443445925905, - "text": "0.019", + "unc_CI_drive": { + "value": [ + -0.0716609947161836, + 0.11220171357678516 + ], + "text": "(-0.072, +0.112)", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank3_experiment": { - "value": "in vivo baseline firing rates of the driven cell classes", - "text": "in vivo baseline firing rates of the driven cell classes", + "unc_S_gain_coef": { + "value": -0.04565580230843003, + "text": "-0.046", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank3_cost": { - "value": "medium", - "text": "medium", + "unc_ST_gain_coef": { + "value": 0.023822223893828185, + "text": "0.024", "unit": null, "note": null, "step": "uncertainty" }, - "voi_rank3_state": { - "value": "null control", - "text": "null control", + "unc_CI_gain_coef": { + "value": [ + -0.10739018244206512, + 0.027653980202641208 + ], + "text": "(-0.107, +0.028)", "unit": null, "note": null, "step": "uncertainty" }, - "voi_unresolved_factors": { - "value": [ - "drive", - "expression", - "gain_coef", - "hill_n", - "lif_seed", - "potency", - "transmitter" - ], - "text": "drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter", + "unc_S_transmitter": { + "value": 0.006078670560203601, + "text": "0.006", "unit": null, "note": null, "step": "uncertainty" }, - "voi_resolved_factors": { - "value": [ - "gain_transform", - "weight_threshold" - ], - "text": "gain_transform, weight_threshold", + "unc_ST_transmitter": { + "value": 0.023642782102253644, + "text": "0.024", "unit": null, "note": null, "step": "uncertainty" }, - "voi_recommendation": { - "value": "Next experiment: calibration of receptor engagement against synaptic gain. Resolving gain_transform would remove about 41% of the model variance of mean_hz for imidacloprid at 1e-06 M (1.29 of 3.15 Hz^2; the readout's SD would fall from 1.77 to 1.36 Hz). the reviewer's central objection: the nicotinic 'buffering' result follows in large part from the deliberately biphasic nAChR rule, so this is the one experiment that could convert a modelling choice into a measurement. Before spending bench time, note that weight_threshold (29%) can be reduced from existing data or compute alone. The remaining factors (drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter) are unresolved at this sample size: their first-order confidence intervals include zero, so this sample cannot rank them and no VOI figure should be quoted for them.", - "text": "Next experiment: calibration of receptor engagement against synaptic gain. Resolving gain_transform would remove about 41% of the model variance of mean_hz for imidacloprid at 1e-06 M (1.29 of 3.15 Hz^2; the readout's SD would fall from 1.77 to 1.36 Hz). the reviewer's central objection: the nicotinic 'buffering' result follows in large part from the deliberately biphasic nAChR rule, so this is the one experiment that could convert a modelling choice into a measurement. Before spending bench time, note that weight_threshold (29%) can be reduced from existing data or compute alone. The remaining factors (drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter) are unresolved at this sample size: their first-order confidence intervals include zero, so this sample cannot rank them and no VOI figure should be quoted for them.", + "unc_CI_transmitter": { + "value": [ + -0.05210207681142619, + 0.06965580923052817 + ], + "text": "(-0.052, +0.070)", "unit": null, "note": null, "step": "uncertainty" }, - "voi_no_experiment_needed": { - "value": [ - "weight_threshold" - ], - "text": "weight_threshold", + "unc_S_hill_n": { + "value": -0.007946568350353221, + "text": "-0.008", "unit": null, "note": null, "step": "uncertainty" }, - "claims_version": { - "value": "1.0", - "text": "1.0", + "unc_ST_hill_n": { + "value": 0.0007525711738826337, + "text": "7.53e-04", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_chain_links": { - "value": 9, - "text": "9", + "unc_CI_hill_n": { + "value": [ + -0.06210277198208717, + 0.04046709008223071 + ], + "text": "(-0.062, +0.040)", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_observed": { - "value": 1, - "text": "1", + "unc_S_potency": { + "value": -0.005989859757710479, + "text": "-0.006", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_literature_derived": { - "value": 1, - "text": "1", + "unc_ST_potency": { + "value": 0.0006784194893530694, + "text": "6.78e-04", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_model_assumption": { - "value": 4, - "text": "4", + "unc_CI_potency": { + "value": [ + -0.06115133451219928, + 0.045925733350771346 + ], + "text": "(-0.061, +0.046)", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_computed": { - "value": 3, - "text": "3", + "unc_S_expression": { + "value": -0.00580912428598648, + "text": "-0.006", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_unknown": { - "value": 0, - "text": "0", + "unc_ST_expression": { + "value": 0.0001280605344204865, + "text": "1.28e-04", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_n_facts": { - "value": 5, - "text": "5", + "unc_CI_expression": { + "value": [ + -0.058783087902346, + 0.04425623004156553 + ], + "text": "(-0.059, +0.044)", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_n_inference": { - "value": 1, - "text": "1", + "unc_S_lif_seed": { + "value": -0.006741189415684353, + "text": "-0.007", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_n_unknown": { - "value": 5, - "text": "5", + "unc_ST_lif_seed": { + "value": 0.0, + "text": "0.000", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_facts_phrase": { - "value": 5, - "text": "5 facts", + "unc_CI_lif_seed": { + "value": [ + -0.06185506580751933, + 0.04343709316115891 + ], + "text": "(-0.062, +0.043)", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_inference_phrase": { - "value": 1, - "text": "1 model inference", + "unc_top_factor": { + "value": "gain_transform", + "text": "gain_transform", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_unknown_phrase": { - "value": 5, - "text": "5 unknowns", + "unc_second_factor": { + "value": "weight_threshold", + "text": "weight_threshold", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_observed_phrase": { - "value": 1, - "text": "1 link", + "unc_noise_floor": { + "value": 0.04565580230843003, + "text": "0.046", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "claims_observed_steps": { - "value": [ - "malecns_edges" - ], - "text": "malecns_edges", + "unc_noise_floor_factor": { + "value": "gain_coef", + "text": "gain_coef", "unit": null, "note": null, - "step": "claims" + "step": "uncertainty" }, - "flylab_version": { - "value": "0.6.0", - "text": "0.6.0", + "unc_noise_floor_definition": { + "value": "the largest magnitude among the negative first-order estimates, which is a measured lower bound on the estimator's error because a true first-order index cannot be negative; it is compared against the null factor's own index and the larger is used", + "text": "the largest magnitude among the negative first-order estimates, which is a measured lower bound on the estimator's error because a true first-order index cannot be negative; it is compared against the null factor's own index and the larger is used", "unit": null, "note": null, - "step": "provenance" + "step": "uncertainty" }, - "git_sha": { - "value": "f95fd88a2165e66d5203db6c6fa8c6a020c5a8e0", - "text": "f95fd88a2165", + "unc_null_factor_S": { + "value": -0.006741189415684353, + "text": "-0.007", "unit": null, "note": null, - "step": "provenance" + "step": "uncertainty" }, - "library_sha256": { - "value": "594b8d32c5e61ebd4de31cc77441aeb82503a0275a7a6ef951ec7a86a48732a8", - "text": "594b8d32c5e6", + "unc_resolved": { + "value": [ + "gain_transform", + "weight_threshold" + ], + "text": "gain_transform, weight_threshold", "unit": null, "note": null, - "step": "provenance" + "step": "uncertainty" }, - "map_id": { - "value": "male-cns:v1.0", - "text": "male-cns:v1.0", + "unc_n_resolved": { + "value": 2, + "text": "2", "unit": null, "note": null, - "step": "provenance" + "step": "uncertainty" }, - "named_nodes": { - "value": 1126, - "text": "1126", + "unc_at_noise_floor": { + "value": [ + "drive", + "expression", + "gain_coef", + "hill_n", + "lif_seed", + "potency", + "transmitter" + ], + "text": "drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter", "unit": null, "note": null, - "step": "provenance" + "step": "uncertainty" }, - "named_edges": { - "value": 1360, - "text": "1360", + "unc_null_control_factors": { + "value": [ + "drive", + "expression", + "gain_coef", + "hill_n", + "lif_seed", + "potency", + "transmitter" + ], + "text": "drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter", "unit": null, "note": null, - "step": "provenance" + "step": "uncertainty" }, - "taste_nodes": { - "value": 1841, - "text": "1841", + "unc_resolution_statement": { + "value": "first-order confidence intervals exclude zero for gain_transform (S1 +0.410, CI [+0.313, +0.500]), weight_threshold (S1 +0.294, CI [+0.199, +0.391]); every other effect is unresolved at this sample size (drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter). 'Unresolved' means the sample cannot separate the effect from zero; it is not a claim that the effect is zero, and a negative point estimate is estimator noise, not a negative contribution. Indistinguishable from the estimator's own noise (floor +0.046, set by gain_coef; the deliberately null factor lif_seed estimates -0.007): drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter.", + "text": "first-order confidence intervals exclude zero for gain_transform (S1 +0.410, CI [+0.313, +0.500]), weight_threshold (S1 +0.294, CI [+0.199, +0.391]); every other effect is unresolved at this sample size (drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter). 'Unresolved' means the sample cannot separate the effect from zero; it is not a claim that the effect is zero, and a negative point estimate is estimator noise, not a negative contribution. Indistinguishable from the estimator's own noise (floor +0.046, set by gain_coef; the deliberately null factor lif_seed estimates -0.007): drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter.", "unit": null, "note": null, - "step": "provenance" + "step": "uncertainty" }, - "taste_edges": { - "value": 19066, - "text": "19\u202f066", + "unc_max_delta_first": { + "value": 0.04034953984299405, + "text": "0.040", "unit": null, "note": null, - "step": "provenance" + "step": "uncertainty" }, - "named_min_weight": { - "value": 5, - "text": "5", + "ish_n_base": { + "value": 16384, + "text": "16\u202f384", "unit": null, "note": null, - "step": "provenance" + "step": "uncertainty" }, - "taste_closure_min_weight": { - "value": 10, - "text": "10", + "ish_estimated": { + "value": [ + 0.3163, + 0.4495, + 0.0072 + ], + "text": "0.3163, 0.4495, 0.0072", "unit": null, "note": null, - "step": "provenance" + "step": "uncertainty" }, - "scale_ladder_sizes": { + "ish_analytic": { "value": [ - 1000, - 5000, - 10000, - 25000, - 50000 + 0.3139, + 0.4424, + 0.0 ], - "text": "1k, 5k, 10k, 25k, 50k", + "text": "0.3139, 0.4424, 0.0000", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_rungs_on_disk": { - "value": [ - "scale_1k", - "scale_5k", - "scale_10k", - "scale_25k", - "scale_50k" - ], - "text": "scale_1k, scale_5k, scale_10k, scale_25k, scale_50k", + "ish_max_abs_error": { + "value": 0.007165317322344709, + "text": "0.007", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_rungs_missing": { - "value": [], - "text": "none", - "unit": null, + "voi_var_total": { + "value": 3.149230029344893, + "text": "3.149", + "unit": "rate-model units^2", "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_n_rungs_on_disk": { - "value": 5, - "text": "5", + "voi_rank1_factor": { + "value": "gain_transform", + "text": "gain_transform", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_1k_name": { - "value": "scale_1k", - "text": "scale_1k", + "voi_rank1_var": { + "value": 1.290081356802269, + "text": "1.290", + "unit": "rate-model units^2", + "note": null, + "step": "uncertainty" + }, + "voi_rank1_fraction": { + "value": 0.4096497698742678, + "text": "0.410", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_1k_nodes": { - "value": 1000, - "text": "1000", + "voi_rank1_experiment": { + "value": "calibration of receptor engagement against synaptic gain", + "text": "calibration of receptor engagement against synaptic gain", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_1k_edges": { - "value": 22857, - "text": "22\u202f857", + "voi_rank1_cost": { + "value": "high", + "text": "high", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_1k_mean_degree": { - "value": 22.857, - "text": "22.9", + "voi_rank1_state": { + "value": "resolved", + "text": "resolved", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_1k_share_onto_seeds": { - "value": 0.06501290633066456, - "text": "7%", + "voi_rank2_factor": { + "value": "weight_threshold", + "text": "weight_threshold", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_1k_in_star": { - "value": false, - "text": "no", + "voi_rank2_var": { + "value": 0.9245370079884832, + "text": "0.925", + "unit": "rate-model units^2", + "note": null, + "step": "uncertainty" + }, + "voi_rank2_fraction": { + "value": 0.29357557224259245, + "text": "0.294", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_frontier_n_paper": { - "value": 1000, - "text": "1000", + "voi_rank2_experiment": { + "value": "no experiment: a reconstruction-confidence analysis", + "text": "no experiment: a reconstruction-confidence analysis", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_largest_feasible_profile": { - "value": "scale_50k", - "text": "scale_50k", + "voi_rank2_cost": { + "value": "none (compute only)", + "text": "none (compute only)", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_largest_feasible_lif": { - "value": "scale_50k", - "text": "scale_50k", + "voi_rank2_state": { + "value": "resolved", + "text": "resolved", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_top_rung": { - "value": "50k", - "text": "50k", + "voi_rank3_factor": { + "value": "drive", + "text": "drive", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_top_profile_hours": { - "value": 21.13708841666667, - "text": "21.1", + "voi_rank3_var": { + "value": 0.05843535970772074, + "text": "0.058", + "unit": "rate-model units^2", + "note": null, + "step": "uncertainty" + }, + "voi_rank3_fraction": { + "value": 0.018555443445925905, + "text": "0.019", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_top_landscape_days": { - "value": 7.397980945833334, - "text": "7", + "voi_rank3_experiment": { + "value": "in vivo baseline firing rates of the driven cell classes", + "text": "in vivo baseline firing rates of the driven cell classes", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_whole_cns_profile_hours": { - "value": 59.427331277777775, - "text": "59", + "voi_rank3_cost": { + "value": "medium", + "text": "medium", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_study_source": { - "value": "papers/scale_study.json", - "text": "papers/scale_study.json", + "voi_rank3_state": { + "value": "null control", + "text": "null control", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_study_status": { - "value": "reported", - "text": "reported", + "voi_unresolved_factors": { + "value": [ + "drive", + "expression", + "gain_coef", + "hill_n", + "lif_seed", + "potency", + "transmitter" + ], + "text": "drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_study_cuts": { + "voi_resolved_factors": { "value": [ - "scale_1k", - "named", - "taste_motor", - "scale_5k", - "scale_10k", - "scale_25k", - "scale_50k" + "gain_transform", + "weight_threshold" ], - "text": "scale_1k, named, taste_motor, scale_5k, scale_10k, scale_25k, scale_50k", + "text": "gain_transform, weight_threshold", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_study_n_cuts": { - "value": 7, - "text": "7", + "voi_recommendation": { + "value": "Next experiment: calibration of receptor engagement against synaptic gain. Resolving gain_transform would remove about 41% of the model variance of mean_hz for imidacloprid at 1e-06 M (1.29 of 3.15 Hz^2; the readout's SD would fall from 1.77 to 1.36 Hz). the reviewer's central objection: the nicotinic 'buffering' result follows in large part from the deliberately biphasic nAChR rule, so this is the one experiment that could convert a modelling choice into a measurement. Before spending bench time, note that weight_threshold (29%) can be reduced from existing data or compute alone. The remaining factors (drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter) are unresolved at this sample size: their first-order confidence intervals include zero, so this sample cannot rank them and no VOI figure should be quoted for them.", + "text": "Next experiment: calibration of receptor engagement against synaptic gain. Resolving gain_transform would remove about 41% of the model variance of mean_hz for imidacloprid at 1e-06 M (1.29 of 3.15 Hz^2; the readout's SD would fall from 1.77 to 1.36 Hz). the reviewer's central objection: the nicotinic 'buffering' result follows in large part from the deliberately biphasic nAChR rule, so this is the one experiment that could convert a modelling choice into a measurement. Before spending bench time, note that weight_threshold (29%) can be reduced from existing data or compute alone. The remaining factors (drive, expression, gain_coef, hill_n, lif_seed, potency, transmitter) are unresolved at this sample size: their first-order confidence intervals include zero, so this sample cannot rank them and no VOI figure should be quoted for them.", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_study_compounds": { + "voi_no_experiment_needed": { "value": [ - "imidacloprid", - "fipronil" + "weight_threshold" ], - "text": "imidacloprid, fipronil", + "text": "weight_threshold", "unit": null, "note": null, - "step": "scale" + "step": "uncertainty" }, - "scale_study_settled": { - "value": true, - "text": "yes", + "claims_version": { + "value": "1.0", + "text": "1.0", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_settled_clean_ladder": { - "value": true, - "text": "yes", + "claims_chain_links": { + "value": 9, + "text": "9", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_statement": { - "value": [ - "imidacloprid: the verdict is 'topology-dependent' on the largest 5 of 7 cuts tested, first reached at 'taste_motor'.", - "fipronil: the verdict is 'topology-dependent' on the largest 7 of 7 cuts tested, first reached at 'scale_1k'." - ], - "text": "imidacloprid: the verdict is 'topology-dependent' on the largest 5 of 7 cuts tested, first reached at 'taste_motor'. fipronil: the verdict is 'topology-dependent' on the largest 7 of 7 cuts tested, first reached at 'scale_1k'.", + "claims_observed": { + "value": 1, + "text": "1", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_sequences": { - "value": [ - "imidacloprid: scale_1k (topology-dependent) -> scale_5k (topology-dependent) -> scale_10k (topology-dependent) -> scale_25k (topology-dependent) -> scale_50k (topology-dependent)", - "fipronil: scale_1k (topology-dependent) -> scale_5k (topology-dependent) -> scale_10k (topology-dependent) -> scale_25k (topology-dependent) -> scale_50k (topology-dependent)" - ], - "text": "imidacloprid: scale_1k (topology-dependent) -> scale_5k (topology-dependent) -> scale_10k (topology-dependent) -> scale_25k (topology-dependent) -> scale_50k (topology-dependent); fipronil: scale_1k (topology-dependent) -> scale_5k (topology-dependent) -> scale_10k (topology-dependent) -> scale_25k (topology-dependent) -> scale_50k (topology-dependent)", + "claims_literature_derived": { + "value": 1, + "text": "1", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_n_range": { - "value": [ - 20, - 1000 - ], - "text": "20-1000", + "claims_model_assumption": { + "value": 4, + "text": "4", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_resolution_range": { - "value": [ - 0.000999000999000999, - 0.047619047619047616 - ], - "text": "0.0010-0.0476", + "claims_computed": { + "value": 3, + "text": "3", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_composition_dominated_cuts": { - "value": [ - "named" - ], - "text": "named", + "claims_unknown": { + "value": 0, + "text": "0", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_n_composition_dominated": { - "value": 1, - "text": "1", + "claims_n_facts": { + "value": 5, + "text": "5", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_every_null_cuts": { - "value": [ - "scale_5k", - "scale_10k" - ], - "text": "scale_5k, scale_10k", + "claims_n_inference": { + "value": 1, + "text": "1", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_coarse_cuts": { - "value": [ - "scale_10k", - "scale_25k", - "scale_50k" - ], - "text": "scale_10k, scale_25k, scale_50k", + "claims_n_unknown": { + "value": 5, + "text": "5", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_coarse_note": { - "value": { - "cuts": [ - "scale_10k", - "scale_25k", - "scale_50k" - ], - "alpha": 0.05 - }, - "text": "at scale_10k, scale_25k, scale_50k the permutation budget falls to n = 20, where the smallest attainable probability is 0.048 against alpha = 0.05: a rejection there is the smallest the test can express", + "claims_facts_phrase": { + "value": 5, + "text": "5 facts", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_imidacloprid_sequence": { - "value": [ - { - "cut": "scale_1k", - "n_nodes": 1000, - "class": "topology-dependent" - }, - { - "cut": "named", - "n_nodes": 1126, - "class": "composition-dominated" - }, - { - "cut": "taste_motor", - "n_nodes": 1841, - "class": "topology-dependent" - }, - { - "cut": "scale_5k", - "n_nodes": 5000, - "class": "topology-dependent" - }, - { - "cut": "scale_10k", - "n_nodes": 10000, - "class": "topology-dependent" - }, - { - "cut": "scale_25k", - "n_nodes": 25000, - "class": "topology-dependent" - }, - { - "cut": "scale_50k", - "n_nodes": 50000, - "class": "topology-dependent" - } - ], - "text": "scale_1k (1000 nodes, mean degree 22.9): topology-dependent; named (1126 nodes, mean degree 1.2): composition-dominated; taste_motor (1841 nodes, mean degree 10.4): topology-dependent; scale_5k (5000 nodes, mean degree 56.0): topology-dependent; scale_10k (10000 nodes, mean degree 62.2): topology-dependent; scale_25k (25000 nodes, mean degree 54.6): topology-dependent; scale_50k (50000 nodes, mean degree 44.3): topology-dependent", + "claims_inference_phrase": { + "value": 1, + "text": "1 model inference", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_imidacloprid_relative_range": { - "value": [ - 0.5198156654684973, - 0.9308633342364611 - ], - "text": "0.52-0.93", + "claims_unknown_phrase": { + "value": 5, + "text": "5 unknowns", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_fipronil_sequence": { - "value": [ - { - "cut": "scale_1k", - "n_nodes": 1000, - "class": "topology-dependent" - }, - { - "cut": "named", - "n_nodes": 1126, - "class": "topology-dependent" - }, - { - "cut": "taste_motor", - "n_nodes": 1841, - "class": "topology-dependent" - }, - { - "cut": "scale_5k", - "n_nodes": 5000, - "class": "topology-dependent" - }, - { - "cut": "scale_10k", - "n_nodes": 10000, - "class": "topology-dependent" - }, - { - "cut": "scale_25k", - "n_nodes": 25000, - "class": "topology-dependent" - }, - { - "cut": "scale_50k", - "n_nodes": 50000, - "class": "topology-dependent" - } - ], - "text": "scale_1k (1000 nodes, mean degree 22.9): topology-dependent; named (1126 nodes, mean degree 1.2): topology-dependent; taste_motor (1841 nodes, mean degree 10.4): topology-dependent; scale_5k (5000 nodes, mean degree 56.0): topology-dependent; scale_10k (10000 nodes, mean degree 62.2): topology-dependent; scale_25k (25000 nodes, mean degree 54.6): topology-dependent; scale_50k (50000 nodes, mean degree 44.3): topology-dependent", + "claims_observed_phrase": { + "value": 1, + "text": "1 link", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, - "scale_study_fipronil_relative_range": { + "claims_observed_steps": { "value": [ - 0.1356619639942983, - 0.5749207286979054 + "malecns_edges" ], - "text": "0.14-0.57", - "unit": null, - "note": null, - "step": "scale" - }, - "scale_study_density_not_size": { - "value": { - "in_star": "named", - "dense": "scale_1k" - }, - "text": "the inversion does not track the number of cells: `scale_1k` has 1000 nodes against `named`'s 1126 and a mean degree of 22.9 against 1.21, and it is already topology-dependent", + "text": "malecns_edges", "unit": null, "note": null, - "step": "scale" + "step": "claims" }, "paper_words_body": { - "value": 6971, - "text": "6971", + "value": 5837, + "text": "5837", "unit": "words", "note": null, "step": "paper" }, "paper_words": { - "value": 10424, - "text": "10\u202f424", + "value": 9442, + "text": "9442", "unit": null, "note": null, "step": "paper" @@ -7469,8 +7416,8 @@ "step": "paper" }, "supplement_words": { - "value": 6676, - "text": "6676", + "value": 5844, + "text": "5844", "unit": "words", "note": null, "step": "paper" @@ -7485,12 +7432,12 @@ "_provenance": { "value": { "flylab_version": "0.6.0", - "git_sha": "f95fd88a2165e66d5203db6c6fa8c6a020c5a8e0", - "library_sha256": "594b8d32c5e61ebd4de31cc77441aeb82503a0275a7a6ef951ec7a86a48732a8", + "git_sha": "50f5465b2df028a6cb2b93926cfbd14d7c746c9b", + "library_sha256": "359096467b33242174dcc5d787a6b7fc6d37573f8daf4771efd159b64ca95e05", "library_version": "0.6", "library_schema_version": 3, "notebook_schema_version": "0.3", - "platform": "Linux-6.18.44-fc-v37-x86_64-with-glibc2.39 python-3.11.15", + "platform": "macOS-26.7-x86_64-i386-64bit python-3.11.15", "map_id": "male-cns:v1.0", "map_citation": "MaleCNS v1.0, HHMI Janelia FlyEM / Cambridge / MRC LMB / Google Research. CC-BY 4.0.", "graphs": { @@ -7516,7 +7463,7 @@ }, "text": "", "unit": null, - "note": "carried over from the previous run in this outdir", + "note": "provenance block, not a paper scalar", "step": "provenance" } }, @@ -7524,31 +7471,31 @@ { "name": "F2_dual_scorecard.png", "path": "figures/F2_dual_scorecard.png", - "bytes": 293669, + "bytes": 240894, "step": "scorecard", "figure": "F2_dual_scorecard", - "caption": "Dual scorecard for imidacloprid and fipronil at 1 \u00b5M. Bars are Hill occupancy of the insect target (blue) and its vertebrate counterpart (orange) for each of the five receptor pairs declared in `library.yaml:selectivity_pairs`. Hatched bars rest on a class placeholder: the compound has no sourced EC50 at that pair and the row is inert. Imidacloprid reaches 0.991 at insect nAChR against 0.091 at vertebrate a4b2; fipronil reaches 0.985 at insect RDL against 0.476 at native heteromeric vertebrate GABA-A. Teaching-tier EC50s: these are literature-order values, not measured constants." + "caption": "Dual scorecard for imidacloprid and fipronil at 1 \u00b5M. Bars are Hill-form engagement of the insect target (blue) and its vertebrate counterpart (orange) for each of the five receptor pairs declared in `library.yaml:selectivity_pairs`. Hatched bars rest on a class placeholder: the compound has no sourced EC50 at that pair and the row is inert. Imidacloprid reaches 0.991 at insect nAChR against 0.091 at vertebrate a4b2; fipronil reaches 0.985 at insect RDL against 0.476 at a multi-species vertebrate GABA-related binding-site proxy. These are heterogeneous assay inputs and transfers, not measured fractional occupancy of matched receptors in living animals." }, { "name": "F11_evidence_types.png", "path": "figures/F11_evidence_types.png", - "bytes": 180063, + "bytes": 150655, "step": "evidence", "figure": "F11_evidence_types", - "caption": "What the compound library rests on, row by row. Left: the parameter each cited source actually measured. Middle: the ordered evidence distance between that source and this compound at this receptor in this species. Right: the transformation the type system then permits, which is a function of the two. Green marks the 10 on-target (E0) rows and the 1 of them that may use a physical binding-occupancy model; amber marks the proxy models, where the same Hill expression is computed but reported as a transfer with a warning naming the gap; red marks rows that carry no modellable evidence and return N/A rather than a small number. The type system, not a convention, enforces this: asking for an engagement from an untyped row raises rather than returning a value, and a soundness invariant is checked exhaustively over every row and every entry point." + "caption": "What the compound library rests on, row by row. Left: the parameter each library row declares. Some source-to-value links remain under assay-level review. Middle: the declared evidence distance between that source and this compound at this receptor in this species. Right: the transformation the type system then permits, which is a function of the two. Green marks the 7 on-target (E0) rows and the 1 of them that may use a physical binding-occupancy model; amber marks the proxy models, where the same Hill expression is computed but reported as a transfer with a warning naming the gap; red marks rows that carry no modellable evidence and return N/A rather than a small number. The type system, not a convention, enforces this: asking for an engagement from an untyped row raises rather than returning a value, and the admission invariant is checked over shipped rows and enumerated entry points. This does not verify the biological validity of admitted values." }, { "name": "F3_occupancy_curves.png", "path": "figures/F3_occupancy_curves.png", - "bytes": 399837, + "bytes": 355967, "step": "curves", "figure": "F3_occupancy_curves", - "caption": "Hill occupancy against free concentration for the four-compound panel, one insect target (solid, blue) and its vertebrate counterpart (dashed, orange) per compound. The shaded band is the selectivity window between the two EC50s, annotated in log10 units; the dotted vertical line is the 1 \u00b5M working concentration used elsewhere in the paper. Ivermectin's window is the GluCl/GlyR pair, deltamethrin's the Nav pair. Curves are the teaching library, not measured concentration-response data." + "caption": "Hill-form model engagement against free concentration for the four-compound panel, one insect target (solid, blue) and its vertebrate counterpart (dashed, orange) per compound. The shaded band is the selectivity window between the two EC50s, annotated in log10 units; the dotted vertical line is the 1 \u00b5M working concentration used elsewhere in the paper. Ivermectin's window is the GluCl/GlyR pair, deltamethrin's the Nav pair. Curves are the teaching library, not measured concentration-response data." }, { "name": "F4_taste_motor_graph.png", "path": "figures/F4_taste_motor_graph.png", - "bytes": 1325154, + "bytes": 1361827, "step": "graph", "figure": "F4_taste_motor_graph", "caption": "Left: the committed MaleCNS v1.0 taste-motor cut (1841 cells, 19066 edges; 1 hop around MN9, DNp01 and the labellar GRN types, plus induced partner-partner edges of at least 10 synapses). Node and edge colour is the predicted consensus transmitter; layout is the deterministic superclass-column layout from `flylab.analysis.layout`. Seeds are marked: sweet GRNs (LB3b/c, up triangles), bitter GRNs (LB1a-d, down triangles), DNp01 (squares) and MN9 (stars). Right: the five strongest sweet and five strongest bitter paths of at most 3 synapses from a labellar GRN to MN9, ranked by the absolute product of signed synapse weights; there are 2595 such sweet and 2372 such bitter paths in total. The sweet/bitter assignment of LB types is a working hypothesis, not a MaleCNS annotation." @@ -7556,55 +7503,55 @@ { "name": "F5_bitter_veto.png", "path": "figures/F5_bitter_veto.png", - "bytes": 145962, + "bytes": 132804, "step": "veto", "figure": "F5_bitter_veto", - "caption": "MN9 firing on the map-extracted labellar-GRN path, driven through the real MaleCNS sweet (34) and bitter (38) GRN body IDs at 150 Hz, never by driving MN9 directly. Left: leaky rate engine. Right: Shiu-style LIF engine with a background Poisson drive. In vehicle, adding bitter drive takes MN9 to zero on both engines (veto ratio 0.00). Fipronil at 1 \u00b5M blocks RDL, raises the sugar-only rate and partially relieves the veto (ratio 0.41 rate, 0.16 LIF). Imidacloprid silences MN9 altogether, so its ratio is undefined rather than zero. Absolute rates differ between engines by construction and are not calibrated to any recording." + "caption": "MN9 firing on the map-extracted labellar-GRN path, driven through the real MaleCNS sweet (34) and bitter (38) GRN body IDs at 150 Hz, never by driving MN9 directly. Left: leaky rate engine. Right: Shiu-style LIF engine with a background Poisson drive. In vehicle, adding bitter drive takes MN9 to zero on both engines (veto ratio 0.00). Fipronil at 1 \u00b5M blocks RDL, raises the sugar-only rate and partially relieves the veto (ratio 0.41 rate, 0.16 LIF). Imidacloprid silences MN9 altogether, so its ratio is undefined rather than zero. Outputs differ between engines by construction and are not calibrated to any recording." }, { "name": "F6_dependence_profile.png", "path": "figures/F6_dependence_profile.png", - "bytes": 334549, + "bytes": 282678, "step": "dependence", "figure": "F6_dependence_profile", - "caption": "Histograms are the drug effect (treated minus vehicle on the network mean rate) measured on 1000 degraded copies of the 1-hop MN9/DNp01 `named` cut; the red line is the same contrast on the real cut. Columns run in increasing order of destruction: transmitter labels permuted, synapse weights permuted, degree-preserving double-edge rewiring, Erdos-Renyi. Titles carry the empirical two-sided permutation p and the three-way verdict against the prespecified equivalence margin (delta = 0.332 Hz, 5 % of the vehicle readout): `distinguishable` means the test rejected, `equivalent within tolerance` means the gap from the null median is below the margin, and `indeterminate` means neither -- a failure to reject is not evidence that the degraded graph gives the same effect. Plain label permutation is shown because it is informative, but it moves the weighted excitation/inhibition balance as well as transmitter identity (T23), so it is a joint null and not a rung of the ladder. Imidacloprid is composition-dominated (necessary level: degree_sequence, equivalent_within_tolerance); fipronil is topology-dependent (necessary level: wiring_with_weighted_transmitter_balance)." + "caption": "Histograms are the drug effect (treated minus vehicle on the network mean rate) measured on 1000 degraded copies of the 1-hop MN9/DNp01 `named` cut; the red line is the same contrast on the real cut. Columns run in increasing order of destruction: transmitter labels permuted while holding outgoing-weight shares near their real values, synapse weights permuted, degree-preserving double-edge rewiring, Erd\u0151s\u2013R\u00e9nyi. Insets show the central 99% of the concentrated Erd\u0151s\u2013R\u00e9nyi null distributions. Titles carry the empirical two-sided permutation p and the three-way verdict against the declared point-gap margin (delta = 0.332 model rate units, 5 % of the vehicle readout): `distinguishable` means the test rejected, `within point-gap tolerance` means the gap from the null median is below the margin after non-rejection, and `not resolved` means neither -- no label is a statistical equivalence test. The plain label permutation is documented in T23: it moves weighted excitation/inhibition balance and is a joint null rather than a ladder rung. Imidacloprid is composition-dominated (necessary level: degree_sequence, equivalent_within_tolerance); fipronil is topology-dependent (necessary level: wiring_with_weighted_transmitter_balance)." }, { "name": "F12_dependence_landscape.png", "path": "figures/F12_dependence_landscape.png", - "bytes": 207481, + "bytes": 192958, "step": "dependence", "figure": "F12_dependence_landscape", - "caption": "Every compound of the library at four concentrations, on both committed cuts, classified by the weakest graph model the permutation test could **not** distinguish from the real cut -- never by a model that reproduces the effect. Classes are decided on Benjamini-Hochberg adjusted probabilities across each landscape's structural tests, with a prespecified relative effect floor of 1 % of the vehicle readout. Left, the 1-hop `named` cut: 11 of 84 cells need the wiring pattern, 42 are not distinguishable from any structure-preserving degradation, and 31 do not move the readout at all. Right, the denser `taste_motor` cut: 54 topology-dependent, 0 composition-dominated, 30 no-effect. The composition-dominated majority does not survive the change of substrate, and the `named` cut is an in-star whose degree-preserving rewire is close to the identity (T19)." + "caption": "Every compound of the library at four concentrations, on both committed cuts, classified by the weakest graph model the permutation test could **not** distinguish from the real cut -- never by a model that reproduces the effect. Classes are decided on Benjamini-Hochberg adjusted probabilities across each landscape's structural tests, with a prespecified relative effect floor of 1 % of the vehicle readout. Left, the 1-hop `named` cut: 0 of 84 cells receive the topology-dependent label, 44 receive composition-dominated and 40 fall below the effect floor. Right, the denser `taste_motor` cut: 44 topology-dependent, 0 composition-dominated, 40 no-effect. The composition-dominated majority does not survive the change of substrate, and the `named` cut is an in-star whose degree-preserving rewire is close to the identity (T19). These are descriptive classes formed from BH-adjusted component tests, without a separate cell-level FDR guarantee." }, { "name": "F16_instrument_validation.png", "path": "figures/F16_instrument_validation.png", - "bytes": 203094, + "bytes": 197220, "step": "dependence", "figure": "F16_instrument_validation", - "caption": "The dependence ladder pointed at a graph whose answer is known in advance. A recurrent cholinergic cycle of known strength is planted in a synthetic cut of the same size, density and transmitter composition as the `named` cut; a gain patch that collapses `g_ach` then removes an amplification that exists only while the cycle is intact, so the drug contrast depends on the wiring by construction. Left: the share of independently generated graphs the ladder classifies topology-dependent, against planted strength and permutation count. Right: the recovery experiment itself, with red marking the strengths the ladder recovered. The unplanted control is the empirical false-positive rate (0 of 18). Without this panel, a negative from the ladder would be indistinguishable from an under-powered test." + "caption": "The dependence ladder pointed at a graph whose answer is known in advance. A recurrent cholinergic cycle of known strength is planted in a synthetic 250-node cut with about 1500 background edges; it is not matched to the real extract's size or connectivity. A gain patch that collapses `g_ach` then removes an amplification that exists only while the cycle is intact, so the drug contrast depends on the wiring by construction. Left: the share of independently generated graphs the ladder classifies topology-dependent, against planted strength and permutation count. Right: the recovery experiment itself, with red marking the strengths the ladder recovered. The unplanted control yielded a detection count (0 of 18). This is a limited positive-control check, not a general power estimate for the paper's biological endpoints." }, { "name": "F13_ablation.png", "path": "figures/F13_ablation.png", - "bytes": 185319, + "bytes": 152743, "step": "ablation", "figure": "F13_ablation", - "caption": "Each level predicts the same 21 compounds with one layer of information removed, and is scored by how well it reproduces the full model's *ordering* of the library (levels have different units, so values are not comparable; the correlation is signed, so a perfect inversion scores -1 rather than 1). Crosses mark the historical depression-only level-C rule, which applied a multiplier of at most 1.0 to every transmitter alike and so could not express disinhibition: its poor correlation was structural, not a finding about connectomes, and the conclusion drawn from it is withdrawn. With a direction-aware rule level C rises to 0.961 at 1 uM from 0.242. The composition level's agreement with the full model must be read against the matched reference distribution of T24, not on its own." + "caption": "Each level predicts the same 21 compounds with one layer of information removed, and is scored by rank agreement with the full model's *ordering* of the library (levels have different units, so values are not comparable; the correlation is signed, so a perfect inversion scores -1 rather than 1). Crosses mark the historical depression-only level-C rule, which applied a multiplier of at most 1.0 to every transmitter alike and so could not express disinhibition: its poor correlation follows partly from that restriction. With a direction-aware rule level C rises to 0.968 at 1 uM from 0.161. The short green interval at B shows the 5th\u201395th percentile of the matched one-gain-per-pseudo-compound B-versus-D reference distribution (T24), with its median dashed. The composition level's agreement with the full model must be read against that distribution." }, { "name": "F7_selectivity_landscape.png", "path": "figures/F7_selectivity_landscape.png", - "bytes": 248793, + "bytes": 228209, "step": "landscape", "figure": "F7_selectivity_landscape", - "caption": "Each point is one compound on one committed graph. The x axis is the receptor selectivity index from the library alone; the y axis is the circuit index, the log10 width of the window between the concentration that moves the simulated network mean rate by 50% and the concentration at which the most potent vertebrate target reaches 20% occupancy. Points below the dashed identity line are buffered by the circuit, points above are amplified. The split is by mechanism, not by potency: chlorpyrifos_oxon, ddt, deltamethrin, permethrin (Nav modulators and the AChE inhibitor) amplify, mean gap +0.29 log10 units, while the nAChR-acting set (acetamiprid, acetylcholine, clothianidin, imidacloprid, nicotine, nitenpyram) buffers, mean gap -0.87. Compounds with no circuit index (chlordimeform, dieldrin, fipronil, gaba, ivermectin, picrotoxin, spinosad, sulfoxaflor, thiamethoxam) never reach the 50% threshold on this cut and are omitted." + "caption": "Each point is one compound on one committed graph. The x axis is the receptor selectivity index from the library alone; the y axis is the circuit index, the log10 width of the window between the concentration that moves the simulated network mean rate by 50% and the concentration at which the most potent vertebrate target reaches 20% occupancy. Points below the dashed identity line are buffered by the circuit, points above are amplified. The split is by mechanism, not by potency: ddt, deltamethrin (Nav modulators and the AChE inhibitor) amplify, mean gap +0.19 log10 units, while the nAChR-acting set (acetamiprid, acetylcholine, clothianidin, imidacloprid, nicotine, nitenpyram) buffers, mean gap -0.87. Compounds with no circuit index (chlordimeform, dieldrin, fipronil, gaba, ivermectin, permethrin, picrotoxin, thiamethoxam) never reach the 50% threshold on this cut and are omitted." }, { "name": "F8_dose_response_sensitivity.png", "path": "figures/F8_dose_response_sensitivity.png", - "bytes": 182736, + "bytes": 181391, "step": "dose", "figure": "F8_dose_response_sensitivity", "caption": "Left: the model's own dose-response for imidacloprid on the 1-hop MN9/DNp01 cut. Points are 4 replicates per concentration with log-normal EC50 jitter (sd 0.3 log10 units) and drive jitter; the curve is a 4-parameter Hill fit, and the band is a 200-resample bootstrap CI on the fitted IC50 (34.7 nM). This is the *model's* circuit IC50, not a receptor IC50 and not an animal IC50. Right: one-at-a-time sensitivity of the same readout at 1 \u00b5M, each parameter halved and doubled. The gain-rule coefficient dominates; the EC50 and Hill coefficient have zero span because occupancy is already saturated at this dose." @@ -7612,15 +7559,15 @@ { "name": "F9_genotype_panel.png", "path": "figures/F9_genotype_panel.png", - "bytes": 179353, + "bytes": 161812, "step": "genotype", "figure": "F9_genotype_panel", - "caption": "Target-receptor occupancy at 1 uM for wild type and each published resistance allele that has a sourced fold-shift for that compound. Grey bars are alleles with no sourced number for the compound: they are left identical to wild type and reported, never extrapolated. Left and centre: fipronil and imidacloprid. Right: the specificity check that motivates the per-compound rule. *para* M918T (super-kdr) shifts deltamethrin 100-fold but leaves DDT at 1-fold, and *Rdl* A301S shifts GABA 33.9-fold while leaving fipronil at 1.1-fold. A toggle that shifted a whole receptor uniformly would be wrong in both cases." + "caption": "Model target engagement at 1 uM for wild type and each published resistance allele that has a sourced fold-shift for that compound. Grey bars are alleles with no sourced number for the compound: they are left identical to wild type and reported, never extrapolated. Left and centre: fipronil and imidacloprid. Right: the specificity check that motivates the per-compound rule. *para* M918T (super-kdr) shifts deltamethrin 100-fold but leaves DDT at 1-fold, and *Rdl* A301S shifts GABA 33.9-fold while leaving fipronil at 1.1-fold. A toggle that shifted a whole receptor uniformly would be wrong in both cases." }, { "name": "F10_rank_validation.png", "path": "figures/F10_rank_validation.png", - "bytes": 250459, + "bytes": 249653, "step": "validation", "figure": "F10_rank_validation", "caption": "Left: the model's compound ordering against each published ordering that the teaching library can cover, with Spearman's rho per entry and the number of compounds it orders. 8 entry-assay combinations are evaluable and 16 are skipped because the library cannot cover them; 4 of the evaluable ones order exactly two compounds, where rho can only be +-1, so no bare mean is quoted: over the 4 entries with more than two compounds the mean is 0.756, and as a concordance count the result is 8 concordant against 0 discordant orderings (two-sided binomial p = 0.008). Right: the two inversions the validation module records rather than fixes. Nitenpyram sits near the top of the library's insect nAChR potency ordering and is the least potent neonicotinoid in *Drosophila* whole-animal bioassays, and clothianidin/imidacloprid are ordered the opposite way from the bioassays. Both are receptor-EC50 versus whole-animal-potency differences: the whole-animal number contains uptake, penetration and metabolism, which FlyLab does not model." @@ -7628,23 +7575,15 @@ { "name": "F14_conclusion_stability.png", "path": "figures/F14_conclusion_stability.png", - "bytes": 263855, + "bytes": 173341, "step": "stability", "figure": "F14_conclusion_stability", - "caption": "Each row is a conclusion this paper could state; each column is one prespecified, admissible way of turning receptor engagement into synaptic gain. Green retains the conclusion, red reverses it, amber means the readout does not exist under that specification. The topology rows are decided on 100 permutations per specification and on Benjamini-Hochberg adjusted probabilities across the 100 structural tests of the run; until v0.6.1 the specification never reached the engine those rows were computed on, so their uniformity was guaranteed by construction rather than measured. Retained by every specification: C2_rdl_disinhibition, C3_imidacloprid_topology_not_distinguishable, C4_fipronil_topology_exceeds, C7_map_bitter_veto. The nicotinic suppression conclusion is reversed by every monotone rule, which is how a result that depends on a modelling choice looks when it is tested rather than asserted; and for a conclusion shaped as a failure to reject, the fraction retained is only as strong as the equivalence behind it (T14's `n_equivalent_within_tolerance` against `n_indeterminate`)." - }, - { - "name": "F15_uncertainty_voi.png", - "path": "figures/F15_uncertainty_voi.png", - "bytes": 192844, - "step": "uncertainty", - "figure": "F15_uncertainty_voi", - "caption": "Left: the share of the variance of the neighbourhood mean rate attributable to each assumption, first-order and total-order, from a Saltelli cross-sample with Jansen estimators, with bootstrap first-order confidence intervals. A factor is **resolved** only when its interval excludes zero; the starred rows are not, and a factor's index exceeding the noise floor does not resolve it. The dashed line is that floor, defined as the largest magnitude among the *negative* first-order estimates: a true first-order index cannot be negative, so that magnitude is a measured lower bound on the estimator's error, and on this run it is several times the declared null factor's own draw (0.046 set by gain_coef, against lif_seed at -0.007). Right: the same shares scaled back into the readout's variance and mapped onto the experiment that would resolve each assumption; a negative estimate is clipped to zero for the decision value, because a value of information cannot be negative. Green marks a factor that needs no experiment at all, only a re-analysis of data already held." + "caption": "Each row is a model-output predicate; each column is one tested way of turning receptor engagement into synaptic gain, with transmitter signs and synapse assignment fixed. Green retains the predicate, red reverses it, and amber means the readout does not exist under that specification. The topology rows are decided on 100 permutations per specification and on Benjamini-Hochberg adjusted probabilities across the 100 structural component tests of the run; this is not a cell-level FDR guarantee. Retained by every tested specification: C2_rdl_disinhibition, C3_imidacloprid_topology_not_distinguishable, C4_fipronil_topology_exceeds, C7_map_bitter_veto. The nicotinic suppression predicate reverses under every monotone rule in this family. For a predicate based on non-rejection, T14 separates descriptive point-gap passes from indeterminate cases; neither is a formal equivalence result." }, { "name": "F1_architecture.png", "path": "figures/F1_architecture.png", - "bytes": 288705, + "bytes": 314341, "step": "architecture", "figure": "F1_architecture", "caption": "FlyLab dataflow. A compound and a free concentration enter the occupancy engine, which reads the sourced library (optionally shifted by a genotype, combined with a mixture partner, or driven by an exposure profile). Occupancy splits into two scorecards. The insect side passes through the mechanism table to a gain patch on transmitter classes of a named MaleCNS cut; the vertebrate side is scored at the same dose and never touches a circuit. The circuit runtime (rate or LIF) produces named-cell readouts, which the analysis and validation layers turn into empirical permutation probabilities with their three-way verdicts, selectivity indices, rank correlations and prospective predictions. Everything lands in one notebook JSON with a provenance block." @@ -7652,10 +7591,18 @@ { "name": "F1_architecture.svg", "path": "figures/F1_architecture.svg", - "bytes": 128097, + "bytes": 128235, "step": "architecture", "figure": "F1_architecture", "caption": null + }, + { + "name": "F15_uncertainty_voi.png", + "path": "figures/F15_uncertainty_voi.png", + "bytes": 163556, + "step": "uncertainty", + "figure": "F15_uncertainty_voi", + "caption": "Left: the share of the variance of the neighbourhood mean rate attributable to each assumption, first-order and total-order, from a Saltelli cross-sample with Jansen estimators, with bootstrap first-order confidence intervals. A factor is **resolved** only when its interval excludes zero; the starred rows are not, and a factor's index exceeding the noise floor does not resolve it. The dashed line is that floor, defined as the largest magnitude among the *negative* first-order estimates: a true first-order index cannot be negative, so that magnitude is a measured lower bound on the estimator's error, and on this run it is several times the declared null factor's own draw (0.046 set by gain_coef, against lif_seed at -0.007). Right: the same shares scaled back into the readout's variance and mapped onto the experiment that would resolve each assumption; a negative estimate is clipped to zero for the heuristic score. This is not formal expected value of information. Green marks a factor that needs no experiment at all, only a re-analysis of data already held." } ], "tables": [ @@ -7676,14 +7623,14 @@ { "name": "T1_library.csv", "path": "tables/T1_library.csv", - "bytes": 32505, + "bytes": 31128, "rows": 102, "step": "scorecard" }, { "name": "T1_library.md", "path": "tables/T1_library.md", - "bytes": 10470, + "bytes": 10386, "rows": 102, "step": "scorecard" }, @@ -7704,14 +7651,14 @@ { "name": "T10_evidence_types.csv", "path": "tables/T10_evidence_types.csv", - "bytes": 29136, + "bytes": 27709, "rows": 102, "step": "evidence" }, { "name": "T10_evidence_types.md", "path": "tables/T10_evidence_types.md", - "bytes": 8412, + "bytes": 8304, "rows": 102, "step": "evidence" }, @@ -7781,7 +7728,7 @@ { "name": "T6_dependence_profile.md", "path": "tables/T6_dependence_profile.md", - "bytes": 2682, + "bytes": 2681, "rows": 15, "step": "dependence" }, @@ -7823,35 +7770,35 @@ { "name": "T22_instrument_validation.md", "path": "tables/T22_instrument_validation.md", - "bytes": 1483, + "bytes": 1571, "rows": 16, "step": "dependence" }, { "name": "T12_dependence_landscape.csv", "path": "tables/T12_dependence_landscape.csv", - "bytes": 31353, + "bytes": 28661, "rows": 84, "step": "dependence" }, { "name": "T12_dependence_landscape.md", "path": "tables/T12_dependence_landscape.md", - "bytes": 9982, + "bytes": 9795, "rows": 84, "step": "dependence" }, { "name": "T12b_dependence_landscape_taste_motor.csv", "path": "tables/T12b_dependence_landscape_taste_motor.csv", - "bytes": 31625, + "bytes": 29345, "rows": 84, "step": "dependence" }, { "name": "T12b_dependence_landscape_taste_motor.md", "path": "tables/T12b_dependence_landscape_taste_motor.md", - "bytes": 10227, + "bytes": 9811, "rows": 84, "step": "dependence" }, @@ -7865,63 +7812,63 @@ { "name": "T24_composition_reference.md", "path": "tables/T24_composition_reference.md", - "bytes": 984, + "bytes": 1015, "rows": 5, "step": "ablation" }, { "name": "T25_engine_normalisation.csv", "path": "tables/T25_engine_normalisation.csv", - "bytes": 1121, + "bytes": 1126, "rows": 14, "step": "ablation" }, { "name": "T25_engine_normalisation.md", "path": "tables/T25_engine_normalisation.md", - "bytes": 1565, + "bytes": 1566, "rows": 14, "step": "ablation" }, { "name": "T13_ablation.csv", "path": "tables/T13_ablation.csv", - "bytes": 11870, + "bytes": 10744, "rows": 84, "step": "ablation" }, { "name": "T13_ablation.md", "path": "tables/T13_ablation.md", - "bytes": 7570, + "bytes": 7569, "rows": 84, "step": "ablation" }, { "name": "T13b_ablation_information.csv", "path": "tables/T13b_ablation_information.csv", - "bytes": 2322, + "bytes": 2346, "rows": 20, "step": "ablation" }, { "name": "T13b_ablation_information.md", "path": "tables/T13b_ablation_information.md", - "bytes": 2318, + "bytes": 2319, "rows": 20, "step": "ablation" }, { "name": "T8_landscape.csv", "path": "tables/T8_landscape.csv", - "bytes": 6960, + "bytes": 7165, "rows": 42, "step": "landscape" }, { "name": "T8_landscape.md", "path": "tables/T8_landscape.md", - "bytes": 3106, + "bytes": 3064, "rows": 42, "step": "landscape" }, @@ -7970,7 +7917,7 @@ { "name": "T5_mixtures.csv", "path": "tables/T5_mixtures.csv", - "bytes": 1097, + "bytes": 1100, "rows": 3, "step": "mixtures" }, @@ -7984,21 +7931,21 @@ { "name": "T14b_specification_diagnostics.csv", "path": "tables/T14b_specification_diagnostics.csv", - "bytes": 2183, + "bytes": 2178, "rows": 25, "step": "stability" }, { "name": "T14b_specification_diagnostics.md", "path": "tables/T14b_specification_diagnostics.md", - "bytes": 2071, + "bytes": 2073, "rows": 25, "step": "stability" }, { "name": "T14_conclusion_stability.csv", "path": "tables/T14_conclusion_stability.csv", - "bytes": 2678, + "bytes": 2699, "rows": 7, "step": "stability" }, @@ -8012,7 +7959,7 @@ { "name": "T15_threshold_grid.csv", "path": "tables/T15_threshold_grid.csv", - "bytes": 752, + "bytes": 753, "rows": 9, "step": "stability" }, @@ -8023,10 +7970,38 @@ "rows": 9, "step": "stability" }, + { + "name": "T26_scale_frontier.csv", + "path": "tables/T26_scale_frontier.csv", + "bytes": 699, + "rows": 9, + "step": "scale" + }, + { + "name": "T26_scale_frontier.md", + "path": "tables/T26_scale_frontier.md", + "bytes": 1285, + "rows": 9, + "step": "scale" + }, + { + "name": "T27_scale_study.csv", + "path": "tables/T27_scale_study.csv", + "bytes": 4239, + "rows": 14, + "step": "scale" + }, + { + "name": "T27_scale_study.md", + "path": "tables/T27_scale_study.md", + "bytes": 2783, + "rows": 14, + "step": "scale" + }, { "name": "T16_uncertainty_budget.csv", "path": "tables/T16_uncertainty_budget.csv", - "bytes": 1286, + "bytes": 1284, "rows": 10, "step": "uncertainty" }, @@ -8040,88 +8015,60 @@ { "name": "T17_value_of_information.csv", "path": "tables/T17_value_of_information.csv", - "bytes": 2466, + "bytes": 2496, "rows": 9, "step": "uncertainty" }, { "name": "T17_value_of_information.md", "path": "tables/T17_value_of_information.md", - "bytes": 1971, + "bytes": 1964, "rows": 9, "step": "uncertainty" }, { "name": "T18_claim_provenance.csv", "path": "tables/T18_claim_provenance.csv", - "bytes": 3371, + "bytes": 3463, "rows": 9, "step": "claims" }, { "name": "T18_claim_provenance.md", "path": "tables/T18_claim_provenance.md", - "bytes": 2907, + "bytes": 2999, "rows": 9, "step": "claims" - }, - { - "name": "T26_scale_frontier.csv", - "path": "tables/T26_scale_frontier.csv", - "bytes": 699, - "rows": 9, - "step": "scale" - }, - { - "name": "T26_scale_frontier.md", - "path": "tables/T26_scale_frontier.md", - "bytes": 1285, - "rows": 9, - "step": "scale" - }, - { - "name": "T27_scale_study.csv", - "path": "tables/T27_scale_study.csv", - "bytes": 4239, - "rows": 14, - "step": "scale" - }, - { - "name": "T27_scale_study.md", - "path": "tables/T27_scale_study.md", - "bytes": 2783, - "rows": 14, - "step": "scale" } ], "notes": [ "[mechanisms] 5 of the 7 transmitter sign magnitudes the rate engine iterates (glutamate -0.4, histamine -0.5, dopamine +0.2, serotonin +0.2, octopamine +0.2) are asserted rather than measured, are exempt from the specification family and are not bounded by the label-permuting null. They are now published in T21 rather than left in the source.", - "[graph] the `named` cut is an in-star: 85% of its 1360 edges terminate on its 4 seed cells, only 184 of 1126 nodes receive any input and only 25% of edges leave a node that itself receives input (mean degree 1.21). A degree-preserving rewire of such a graph is close to the identity, so a negative topology verdict on this cut is weak evidence. The `taste_motor` cut (mean degree 10.36) is the substrate to generalise from.", + "[graph] the `named` cut is an in-star: 85% of its 1360 edges terminate on its 4 seed cells, only 184 of 1126 nodes receive any input and only 25% of edges leave a node that itself receives input (mean degree 1.21). A degree-preserving rewire of such a graph is close to the identity, so a negative topology verdict on this cut is weak evidence. The `taste_motor` cut (mean degree 10.36) is the denser comparison substrate.", "[graph] the `named` cut contains no octopaminergic cell at all, so a compound whose only insect mechanism acts through g_oct (chlordimeform) cannot move this readout for substrate reasons rather than pharmacological ones.", - "[veto] imidacloprid at 1 uM silences MN9 on the map path on both engines (sugar-driven MN9 = 0 Hz), so its bitter-veto ratio is undefined, not 0. g_ach hits the 0.05 floor, which removes the cholinergic GRN->MN9 drive itself. This is a property of the agonist patch rule, and it is why the taste-map arm of the null panel is undefined for imidacloprid.", - "[dependence] the transmitter null matters to the equivalence claim and not to the probability: imidacloprid's gap from the null median is 2.65 Hz under the plain label permutation and 0.24 Hz under the weight-matched null, against a prespecified margin of 0.332 Hz. The plain permutation is a joint target-set-and-sign null and is reported as one.", - "[dependence] the fipronil bitter-veto ratio on the taste-motor path is a negative at n=300 shuffles (max |z| = 0.220, p resolution 0.0033): the veto ratio is model behaviour, not evidence about the wiring, and the verdicts say which modes are equivalent within tolerance (none) and which are merely indeterminate (erdos_renyi, rewire_degree_preserving, sign_permute, sign_permute_weight_matched, weight_permute).", + "[veto] imidacloprid at 1 uM silences MN9 on the map path on both engines (sugar-driven MN9 = 0 in each engine's output units), so its bitter-veto ratio is undefined, not 0. g_ach hits the 0.05 floor, which removes the cholinergic GRN->MN9 drive itself. This is a property of the agonist patch rule, and it is why the taste-map arm of the null panel is undefined for imidacloprid.", + "[dependence] the transmitter null changes the descriptive point-gap verdict: imidacloprid's gap from the null median is 2.65 rate-model units under the plain label permutation and 0.24 under the weight-matched null, against a declared margin of 0.332 rate-model units. The plain permutation is a joint target-set-and-sign null and is reported as one.", + "[dependence] the fipronil bitter-veto ratio on the taste-motor path is a negative at n=300 shuffles (max |z| = 0.220, p resolution 0.0033): the veto ratio is model behaviour, not evidence about the wiring, and the verdicts say which modes pass the descriptive point-gap rule (none) and which are merely indeterminate (erdos_renyi, rewire_degree_preserving, sign_permute, sign_permute_weight_matched, weight_permute).", "[dependence] plain transmitter-label permutation does NOT preserve the weighted excitation/inhibition balance: the cholinergic share of total outgoing synaptic weight is 0.619 on the real `named` cut against 0.544 +- 0.027 over 1000 permutations, putting the real graph at the 100th percentile of its own null. The weight-matched null holds every tracked share within 0.002.", - "[dependence] instrument validation: the ladder recovered 2 of 3 planted topology-dependent effects at n = 200 and returned no false positive on the unplanted control. Detection over the planted grid by permutation count is n=50: 0.61, n=200: 0.61, n=1000: 0.67 and by planted strength strength 0.25: 0.00-0.17; strength 0.5: 0.67-0.83; strength 1: 1.00-1.00; the empirical false-positive rate on the unplanted control is 0 of 18. Effect size dominates the permutation count: full power at planted strength 1, detection at or below one in four at 0.25.", - "[dependence] multiplicity changes the headline count on the `named` cut: 14 cells are topology-dependent on the raw permutation p and 11 survive Benjamini-Hochberg across the 168 structural tests of the landscape. The corrected count is the one the paper quotes.", - "[dependence] the central RQ2 result REVERSES on a cut that has topology: repeated on `taste_motor` (1841 nodes, 19066 edges, mean degree 10.4) the landscape returns 0 composition-dominated cells of 84, against 42 of 84 on the in-star `named` cut. The composition-dominated majority is a property of the substrate, not of receptor perturbation on connectomes.", + "[dependence] instrument validation: the ladder recovered 2 of 3 planted topology-dependent effects at n = 200 and returned no false positive on the unplanted control. Detection over the planted grid by permutation count is n=50: 0.61, n=200: 0.61, n=1000: 0.67 and by planted strength strength 0.25: 0.00-0.17; strength 0.5: 0.67-0.83; strength 1: 1.00-1.00; unplanted controls were detected in 0 of 18 runs. In this finite grid, observed detection was 1.00 at planted strength 1 and at most 0.25 at 0.25; these fractions are not general power estimates.", + "[dependence] multiplicity changes the headline count on the `named` cut: 11 cells are topology-dependent on the raw permutation p and 0 survive Benjamini-Hochberg across the 168 structural tests of the landscape. The corrected count is the one the paper quotes.", + "[dependence] the descriptive class distribution differs on a denser cut: repeated on `taste_motor` (1841 nodes, 19066 edges, mean degree 10.4) the landscape returns 0 composition-dominated cells of 84, against 44 of 84 on the in-star `named` cut. The composition-dominated majority is a property of the tested cut and runtime, not a general result about connectomes.", "[ablation] the ablation ladder's ranking is concentration-dependent: with the historical depression-only level-C rule the weakest level is 10 nM: C_topology_only; 100 nM: A_receptor_only; 1 \u00b5M: A_receptor_only; 10 \u00b5M: A_receptor_only. Any sentence of the form 'the topology-only level is the worst of the four' is true of one column and false of the others.", - "[ablation] the level-C conclusion is withdrawn: with a direction-aware generic rule the topology-only baseline's rank correlation with the full model at 1 uM rises from 0.242 (depression-only floor, which cannot express disinhibition and whose every entry is at most zero) to 0.961. 'The connectome without the pharmacology is not a cheap substitute' was a statement about a sign-broken baseline.", - "[ablation] the composition-versus-full correlation carries no compound-level information: shuffling the compound-to-gain assignment leaves it identically 0.988, and pharmacology-free pseudo-compounds whose gain vectors have the shape the mechanism rules produce already reach a median of 0.847. Both levels are functions of the same gain vector, so the number is a statement about the map from gains to readouts.", - "[ablation] the composition verdict is not robust to the engine's row normalisation: the B-versus-D rank correlation at 1 uM is 0.988 under the shipped `row_abs` mode and 0.666 under a degree-corrected one. The normalisation is an undocumented modelling choice that makes each cell's recurrent input a composition-weighted average of its presynaptic gains.", - "[landscape] no circuit selectivity index exists for chlordimeform, dieldrin, fipronil, gaba, ivermectin, picrotoxin, spinosad, sulfoxaflor, thiamethoxam: on this neighbourhood their largest relative change in the mean rate never reaches the 50% threshold (RDL/GluCl blockers are bounded by the GABA/glutamate share of the cut and by the 0.05 gain floor).", - "[dose] at 1 uM imidacloprid the model is exactly insensitive to a 2x change in the insect_nAChR EC50 and in the Hill coefficient (span 0 Hz): occupancy is saturated, so the gain-rule coefficient is the only parameter that still moves the readout. A dose-response experiment, not a single high dose, is what constrains the EC50.", + "[ablation] the level-C rank comparison depends on the generic rule: with a direction-aware rule the topology-only baseline's rank correlation with the full model at 1 uM rises from 0.161 (depression-only floor, which cannot express disinhibition and whose every entry is at most zero) to 0.968. 'The connectome without the pharmacology is not a cheap substitute' was a statement about a sign-broken baseline.", + "[ablation] the composition-versus-full correlation carries no compound-level evidence from a label shuffle: shuffling the compound-to-gain assignment leaves it identically 0.921, and pharmacology-free pseudo-compounds whose gain vectors have the shape the mechanism rules produce already reach a median of 0.847. Both levels are functions of the same gain vector, so the number is a statement about the map from gains to readouts.", + "[ablation] the composition verdict is not robust to the engine's row normalisation: the B-versus-D rank correlation at 1 uM is 0.921 under the shipped `row_abs` mode and 0.551 under a degree-corrected one. The normalisation is an documented modelling choice that makes each cell's recurrent input a composition-weighted average of its presynaptic gains.", + "[landscape] no circuit selectivity index exists for chlordimeform, dieldrin, fipronil, gaba, ivermectin, permethrin, picrotoxin, thiamethoxam: on this neighbourhood their largest relative change in the mean rate never reaches the 50% threshold (RDL/GluCl blockers are bounded by the GABA/glutamate share of the cut and by the 0.05 gain floor).", + "[dose] at 1 uM imidacloprid the model is exactly insensitive to a 2x change in the insect_nAChR EC50 and in the Hill coefficient (span 0 rate-model units): occupancy is saturated, so the gain-rule coefficient is the only parameter that still moves the readout. A dose-response experiment, not a single high dose, is what constrains the EC50.", "[validation] the literature-concordance mean is not quotable bare: 4 of the 8 evaluable comparisons order exactly two compounds, where Spearman's rho can only be +-1 and the two-sided p is 1.0 whatever the model does. Over the 4 comparisons with more than two compounds the mean is 0.756; as a concordance count it is 8 concordant against 0 discordant orderings (two-sided binomial p = 0.008).", - "[stability] the C3 retention rate is no longer vacuous and is no longer uniform: at n_shuffles = 100 (resolution 0.0099, against the previous 6 at which p <= 0.05 was arithmetically unattainable) the conclusion is retained by 25 of 25 specifications, but only 17 of those reach equivalence within the prespecified margin; 8 are merely indeterminate.", - "[stability] the suppression conclusion is specification-dependent: 10 of 25 prespecified specifications reverse it. Under linear@1.00x the same compound at the same engagement EXCITES the network (10.1 Hz treated vs 6.631 Hz vehicle) instead of suppressing it. Suppression is a property of the biphasic desensitisation term, not of the pharmacology-connectome integration.", + "[stability] the C3 retention rate is no longer vacuous and is no longer uniform: at n_shuffles = 100 (resolution 0.0099, against the previous 6 at which p <= 0.05 was arithmetically unattainable) the conclusion is retained by 25 of 25 specifications, but only 17 of those pass the descriptive point-gap rule; 8 are indeterminate.", + "[stability] the suppression conclusion is specification-dependent: 10 of 25 prespecified specifications reverse it. Under linear@1.00x the same compound at the same engagement raises the model readout (10.1 treated vs 6.631 vehicle, rate-model units) instead of suppressing it. Suppression is a property of the biphasic desensitisation term, not of the pharmacology-connectome integration.", "[stability] the supplement's claim that the monotone rules 'buffer at least as much, never less' is false: 3 of the 6 reversing specifications with a defined index buffer LESS than the shipped rule (linear@1.50x, saturating@1.25x, saturating@1.50x), and the aggregate of -0.893 is carried by the extreme of a range running -0.199 to -1.734.", "[stability] T8 and T15 have different scopes and must be read that way: T8 scores the whole library on both cuts, T15 only the 10 compounds of the two mechanism classes the conclusion is about. acetylcholine carries a circuit index in T8 and is outside T15's compound set entirely.", - "[stability] the amplify/buffer split is NOT stable across the 3x3 threshold grid: nav_ache fails at circuit_frac=0.50,vert_limit=0.10. The paper quotes the range over the grid, never the point estimate.", + "[stability] the amplify/buffer split is NOT stable across the 3x3 threshold grid: nav_ache fails at circuit_frac=0.40,vert_limit=0.10, circuit_frac=0.50,vert_limit=0.10. The paper quotes the range over the grid, never the point estimate.", + "[scale] scale study: 7 cuts (scale_1k, named, taste_motor, scale_5k, scale_10k, scale_25k, scale_50k); verdicts settled on the clean ladder: yes. imidacloprid: the verdict is 'topology-dependent' on the largest 5 of 7 cuts tested, first reached at 'taste_motor'. fipronil: the verdict is 'topology-dependent' on the largest 7 of 7 cuts tested, first reached at 'scale_1k'.", "[uncertainty] the estimator's noise floor is 0.046, set by the most negative first-order estimate (gain_coef), not by the declared null factor lif_seed at -0.007. Only gain_transform, weight_threshold have a first-order confidence interval that excludes zero; every other factor is unresolved at this sample size, and a negative point estimate is estimator error rather than a negative contribution.", - "[claims] the provenance chain behind a circuit readout has 9 links and exactly 1 of them (malecns_edges) is a measurement of the system being simulated; the rest are literature reuse, asserted modelling choices and computation.", - "[scale] scale study: 7 cuts (scale_1k, named, taste_motor, scale_5k, scale_10k, scale_25k, scale_50k); verdicts settled on the clean ladder: yes. imidacloprid: the verdict is 'topology-dependent' on the largest 5 of 7 cuts tested, first reached at 'taste_motor'. fipronil: the verdict is 'topology-dependent' on the largest 7 of 7 cuts tested, first reached at 'scale_1k'." + "[claims] the provenance chain behind a circuit readout has 9 links and exactly 1 of them (malecns_edges) is a measurement of the system being simulated; the rest are literature reuse, asserted modelling choices and computation." ], - "partial": false, + "partial": true, "honesty": [ "Every number in this file is model_derived: it comes from the teaching EC50 library, the gain patch rules and one or two hops-limited MaleCNS cuts. Nothing here was measured in a living fly.", "live_lab is null in every notebook this pipeline produced.", diff --git a/papers/tables/T10_evidence_types.csv b/papers/tables/T10_evidence_types.csv index 5d0a955..39b0ba5 100644 --- a/papers/tables/T10_evidence_types.csv +++ b/papers/tables/T10_evidence_types.csv @@ -1,5 +1,5 @@ compound,receptor,param_type,value_M,hill_n,direction,relation,evidence_distance,species,evidence_tier,engagement_model,source -imidacloprid,insect_nAChR,EC50,2e-08,1.2,agonist,exact_compound_related_receptor,E2,Ctenocephalides felis alpha1/alpha2 + Gallus gallus beta2 hybrid,literature_order,functional_engagement_proxy,"Dederer et al. 2011, Insect Biochem Mol Biol 41:51 (Drosophila/Ctenocephalides alpha1-beta1 nAChR; neonicotinoid agonist potency in the 10-100 nM order). Value kept unchanged from FlyLab v0.4. SOURCE DESCRIPTION CORRECTED in v0.6 (schema v3): Dederer et al. expressed cat-flea Cfalpha1/Cfalpha2 with the CHICKEN beta2 subunit, i.e. a hybrid receptor - not an 'alpha1-beta1' receptor as the v0.5 source string said. The value is unchanged; only the description of what was measured is corrected. It is an agonist EC50 on a hybrid receptor, so this row is typed EC50 / exact_compound_related_receptor (evidence distance E2) and drives functional_engagement_proxy, never binding occupancy. The subunit-resolved beta1 row is insect_nAChR_beta1." +imidacloprid,insect_nAChR,EC50,2e-08,1.2,agonist,exact_compound_related_receptor,E2,Ctenocephalides felis alpha1/alpha2 + Gallus gallus beta2 hybrid,literature_order,functional_engagement_proxy,"Dederer H, Werr M, Ilg T (2011) Insect Biochem Mol Biol 41(1):51-61. DOI 10.1016/j.ibmb.2010.09.012; PMID 20933086. Imidacloprid EC50 was 0.02 uM on cat-flea Cfalpha1/chicken Gg-beta2 hybrid receptors. The same paper also studied Drosophila alpha subunits with chicken beta2, but the 0.02 uM value is not a Drosophila receptor measurement. This unchanged library value is therefore an E2 related-receptor functional-engagement proxy, not binding occupancy." imidacloprid,vertebrate_nAChR_a4b2,EC50,1e-05,1.0,agonist,class_extrapolation,E3,mammalian (review-level contrast),literature_order,functional_engagement_proxy,"Tomizawa and Casida 2005, Annu Rev Pharmacol Toxicol 45:247 (imidacloprid selectivity for insect over vertebrate nAChR is >1000-fold; vertebrate potency in the micromolar order). Order-of-magnitude selectivity contrast, kept from FlyLab v0.4." imidacloprid,vertebrate_GABA_A,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder: no neonicotinoid action at GABA-A assumed imidacloprid,insect_RDL,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder: no neonicotinoid action at RDL assumed @@ -19,8 +19,8 @@ diazepam,vertebrate_GABA_A,EC50,1e-08,1.3,positive_modulator,exact_compound_exac diazepam,insect_RDL,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,"class-order placeholder: insect RDL lacks the vertebrate benzodiazepine site, so diazepam is modelled as inactive on the insect GABA-Cl channel." fipronil,insect_nAChR,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder fipronil,vertebrate_nAChR_a4b2,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder -fipronil,vertebrate_GABA_A,IC50,1.1e-06,1.0,antagonist,exact_compound_exact_receptor_other_species,E1,Homo sapiens recombinant GABA-A ([3H]EBOB displacement),literature_order,functional_engagement_proxy,"Ratra and Casida 2001, Toxicol Lett 122:215 ([3H]EBOB displacement, human recombinant GABA-A): vertebrate average IC50 for fipronil is 1103 nM = 1.1e-6 M. CORRECTED in v0.5: FlyLab v0.4 carried 1.0e-05 M while citing this same paper, so the library contradicted its own citation; the value is now the cited one. SUBUNIT CAVEAT: the human beta3 HOMOpentamer is far more sensitive, 0.5-2.4 nM (Ratra, Kamita and Casida 2001, Toxicol Appl Pharmacol 172:233), so the 'vertebrate-safe' reading of this row applies to native heteromeric receptors only, not to beta3 homomers." -fipronil,insect_RDL,IC50,3e-08,1.2,antagonist,exact_compound_exact_receptor_exact_species,E0,Drosophila melanogaster,literature_order,functional_engagement,"Hosie et al. 1995, Br J Pharmacol 115:909 (fipronil block of dieldrin-sensitive Drosophila GABA receptors, tens-of-nM order). SPREAD NOTE: the literature spans ~1500-fold for this one row - 2.4-12 nM on house-fly head membranes ([3H]EBOB displacement; Cole et al. 1993, Pestic Biochem Physiol 46:47 and Ratra and Casida 2001, Toxicol Lett 122:215) versus 0.3-3 uM on the predominant in vivo splice/edit variant of Drosophila RDL in Xenopus oocytes (Lees et al. 2014, PLoS ONE 9:e97468). Library keeps the potent end (unchanged from FlyLab v0.4); it is a defensible midpoint-to-potent choice, not a measured constant." +fipronil,vertebrate_GABA_A,IC50,1.1e-06,1.0,antagonist,exact_compound_exact_receptor_other_species,E1,native GABA-related chloride-channel preparations from multiple vertebrate species ([3H]EBOB displacement),literature_order,functional_engagement_proxy,"Hainzl, Cole and Casida 1998, Chem Res Toxicol 11:1529-1535 (DOI 10.1021/tx980157t; PMID 9860498): mean fipronil IC50 1103 nM across human, dog, mouse, chicken, quail and salmon preparations by [3H]EBOB displacement. This is a multi-species binding-site proxy, not a human recombinant functional GABA-A potency." +fipronil,insect_RDL,IC50,3e-08,1.2,antagonist,exact_compound_related_receptor,E2,insect GABA-related chloride-channel preparations; Drosophila RDL variant response differs by assay,literature_order,functional_engagement_proxy,"The 30 nM model value is an assay-transfer proxy, not a measured constant for the predominant Drosophila RDLbd I/V variant. Hainzl et al. 1998 (DOI 10.1021/tx980157t) reports 3-12 nM insect [3H]EBOB displacement; Lees et al. 2014 (DOI 10.1371/journal.pone.0097468) reports about 0.63 uM functional block on RDLbd I/V. Model sensitivity to this range is unresolved." clothianidin,insect_nAChR,EC50,3e-08,1.2,agonist,exact_compound_related_receptor,E2,Periplaneta americana native neuronal nAChR,literature_order,functional_engagement_proxy,"Thany 2009, Neurotoxicology 30:1045 (clothianidin is an agonist at cockroach nAChRs from about 30 nM); Tomizawa and Casida 2005, Annu Rev Pharmacol Toxicol 45:247 for low-nM insect binding." clothianidin,vertebrate_nAChR_a7,EC50,0.00074,1.0,partial_agonist,exact_compound_exact_receptor_other_species,E1,Rattus norvegicus alpha7,literature_order,functional_engagement_proxy,"Cartereau et al. 2018, Br J Pharmacol 175 (rat alpha7; clothianidin partial agonist, EC50 0.74 mM, vs acetylcholine 0.16 mM)." clothianidin,vertebrate_nAChR_a4b2,EC50,0.0001,1.0,agonist,class_extrapolation,E3,mammalian (inferred from the alpha7 data),literature_order,functional_engagement_proxy,"Order-of-magnitude contrast from the mammalian alpha7 data of Cartereau et al. 2018, Br J Pharmacol 175; no direct alpha4beta2 EC50 was located, so this row is a high-micromolar class order, not a measured EC50." @@ -36,17 +36,17 @@ acetamiprid,vertebrate_nAChR_a7,EC50,0.00073,1.0,partial_agonist,exact_compound_ acetamiprid,vertebrate_nAChR_a4b2,EC50,0.0001,1.0,agonist,class_extrapolation,E3,mammalian (inferred from the alpha7 data),literature_order,functional_engagement_proxy,"Order-of-magnitude contrast from the mammalian alpha7 data of Cartereau et al. 2018, Br J Pharmacol 175; not a measured alpha4beta2 EC50." acetamiprid,insect_RDL,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder acetamiprid,vertebrate_GABA_A,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder -sulfoxaflor,insect_nAChR,EC50,0.00015,1.0,agonist,exact_compound_related_receptor,E2,Periplaneta americana DUM neurons (alpha-bungarotoxin-insensitive nAChR),literature_order,functional_engagement_proxy,"Watson et al. 2011, Pestic Biochem Physiol 101:32 (sulfoxaflor fully activates alpha-bungarotoxin-insensitive nAChRs of cockroach DUM neurons, EC50 about 154 uM); Sparks et al. 2013, Pestic Biochem Physiol 107:1 (mode of action review)." +sulfoxaflor,insect_nAChR,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,"Placeholder pending source verification: Watson et al. 2011, Insect Biochem Mol Biol 41:432-439 (DOI 10.1016/j.ibmb.2011.01.009; PMID 21296156) supports Dalpha2/chicken beta2 receptor activity but does not verify the former 154 uM value in the checked record. A table-verified potency is required." sulfoxaflor,vertebrate_nAChR_a4b2,EC50,0.001,1.0,agonist,class_extrapolation,E3,mammalian (review-level contrast),literature_order,functional_engagement_proxy,"Sparks et al. 2013, Pestic Biochem Physiol 107:1 (low mammalian nAChR activity reported for sulfoximines). Order-of-magnitude contrast, not a measured EC50." sulfoxaflor,vertebrate_nAChR_a7,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder sulfoxaflor,insect_RDL,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder sulfoxaflor,vertebrate_GABA_A,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder -spinosad,insect_nAChR,EC50,5e-06,1.0,positive_modulator,exact_compound_related_receptor,E2,Drosophila melanogaster Dalpha6 / heterologous insect alpha6-like receptors,literature_order,functional_engagement_proxy,"Perry et al. 2015, Insect Biochem Mol Biol 64:106 (Drosophila Dalpha6 is required for spinosad action in vivo); heterologous alpha6 nAChR studies report allosteric activation in the low-micromolar order (EC50 about 5 uM). AGGREGATE ROW (v0.6): the measurement behind this number is alpha6-specific and is recorded subunit-resolved in insect_nAChR_alpha6. The same value is repeated here on the documented aggregate key (no new number was invented) so the legacy g_ach gain rule still fires; read as an aggregate it overstates the receptor population the value applies to." +spinosad,insect_nAChR,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,Placeholder pending source verification: Perry et al. 2015 (DOI 10.1016/j.ibmb.2015.01.017; PMID 25747008) supports Dalpha6 target involvement but not a 5 uM receptor EC50. Martelli et al. 2022 (PMID 35191376) reports low-dose antagonism of Dalpha6-mediated cholinergic response. No quantitative receptor potency or direction is admitted pending an assay-matched source. spinosad,vertebrate_nAChR_a4b2,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder: no mammalian nAChR potency is reported for spinosyns at insecticidal concentrations. spinosad,vertebrate_nAChR_a7,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder spinosad,insect_RDL,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder spinosad,vertebrate_GABA_A,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder -spinosad,insect_nAChR_alpha6,EC50,5e-06,1.0,positive_modulator,exact_compound_related_receptor,E2,Drosophila melanogaster Dalpha6 (in vivo requirement) / heterologous insect alpha6-like receptors (the exact preparation behind the 5 uM order is not pinned),literature_order,functional_engagement_proxy,"Perry et al. 2015, Insect Biochem Mol Biol 64:106 (Drosophila Dalpha6 is required for spinosad action in vivo; Dalpha6 knockouts are highly resistant); heterologous alpha6 nAChR studies report allosteric activation in the low-micromolar order (EC50 about 5 uM). SUBUNIT-RESOLVED ROW (v0.6): this is the assay the spinosad number actually came from. The identical value also appears on the aggregate insect_nAChR key so the legacy gain rules keep working; no new number was invented for the split." +spinosad,insect_nAChR_alpha6,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,"Placeholder pending source verification: Dalpha6 is a spinosad target (Perry et al. 2015, DOI 10.1016/j.ibmb.2015.01.017), but that study does not measure the previously entered 5 uM receptor EC50; Martelli et al. 2022 (PMID 35191376) reports low-dose antagonism. This row is intentionally not modelled until potency and effect direction are sourced." acetylcholine,insect_nAChR,EC50,1e-05,1.2,agonist,exact_compound_related_receptor,E2,Drosophila melanogaster Dalpha1 + Gallus gallus beta2 hybrid (and Dalpha5 homomers),literature_order,functional_engagement_proxy,"Ihara et al. 2018, Br J Pharmacol 175:1999 (Drosophila Dalpha1/chicken beta2 hybrids; acetylcholine reference responses in the low-micromolar range). Drosophila Dalpha5 homomers report an acetylcholine EC50 of about 8.8 uM." acetylcholine,vertebrate_nAChR_a4b2,EC50,1e-05,1.2,agonist,exact_compound_exact_receptor_other_species,E1,Homo sapiens alpha4beta2 in Xenopus oocytes,literature_order,functional_engagement_proxy,"Chavez-Noriega et al. 1997, J Pharmacol Exp Ther 280:346 (human alpha4beta2 in oocytes; acetylcholine EC50 in the low-micromolar range)." acetylcholine,vertebrate_nAChR_a7,EC50,0.00016,1.5,agonist,exact_compound_exact_receptor_other_species,E1,Rattus norvegicus alpha7,literature_order,functional_engagement_proxy,"Cartereau et al. 2018, Br J Pharmacol 175 (rat alpha7; acetylcholine EC50 0.16 mM)." @@ -59,7 +59,7 @@ gaba,insect_nAChR,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modell gaba,vertebrate_nAChR_a4b2,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder picrotoxin,insect_RDL,IC50,1e-06,1.0,antagonist,exact_compound_exact_receptor_exact_species,E0,Drosophila melanogaster,literature_order,functional_engagement,"Hosie et al. 1995, Br J Pharmacol 115:909 (picrotoxinin block of Drosophila GABA receptors, micromolar order); RDL oocyte IC50 varies up to about 17-fold with expression conditions (PLoS ONE 2021, RDL expression-condition study)." picrotoxin,vertebrate_GABA_A,IC50,1e-06,1.0,antagonist,exact_compound_exact_receptor_other_species,E1,mammalian GABA-A,literature_order,functional_engagement_proxy,"Ratra and Casida 2001, Toxicol Lett 122:215 (picrotoxinin blocks mammalian GABA-A at about 1 uM, i.e. no useful insect selectivity)." -picrotoxin,vertebrate_GlyR,IC50,1e-05,1.0,antagonist,class_extrapolation,E3,mammalian homomeric glycine receptor,literature_order,functional_engagement_proxy,"Ratra and Casida 2001, Toxicol Lett 122:215 and standard glycine receptor pharmacology (picrotoxin blocks homomeric GlyR in the micromolar order). Order-of-magnitude value." +picrotoxin,vertebrate_GlyR,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,Placeholder pending source verification: The previously cited Ratra and Casida GABA-receptor paper does not assay GlyR. A direct glycine-receptor picrotoxin preparation and potency are required before this row can be quantified. picrotoxin,insect_nAChR,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder picrotoxin,vertebrate_nAChR_a4b2,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder dieldrin,insect_RDL,IC50,1e-07,1.0,antagonist,exact_compound_exact_receptor_exact_species,E0,Drosophila melanogaster,literature_order,functional_engagement,"ffrench-Constant et al. 1993, Nature 363:449 (Rdl A301S confers cyclodiene resistance); Hosie et al. 1995, Br J Pharmacol 115:909 (cyclodiene block of dieldrin-sensitive Drosophila GABA receptors in the 0.1 uM order)." @@ -67,11 +67,11 @@ dieldrin,vertebrate_GABA_A,IC50,1e-06,1.0,antagonist,exact_compound_exact_recept dieldrin,insect_nAChR,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder dieldrin,vertebrate_nAChR_a4b2,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder ivermectin,insect_GluCl,EC50,2e-08,1.0,agonist,exact_compound_related_receptor,E2,invertebrate GluCl (Drosophila genetics plus heterologous invertebrate GluCl),literature_order,functional_engagement_proxy,"Kane et al. 2000, PNAS 97:13949 (drug-resistant Drosophila show GluCl is the avermectin target); invertebrate GluCl activation is reported in the 0.1-30 nM order." -ivermectin,insect_RDL,EC50,1e-06,1.0,positive_modulator,exact_compound_exact_receptor_exact_species,E0,Drosophila melanogaster RDL (predominant in vivo variant) in Xenopus oocytes,literature_order,functional_engagement,"Lees et al. 2014, PLoS ONE 9:e97468 (ivermectin activates and potentiates the predominant in vivo RDL variant at sub-micromolar to micromolar)." +ivermectin,insect_RDL,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,"Placeholder pending source verification: Lees et al. 2014 (DOI 10.1371/journal.pone.0097468) reports ivermectin inhibition of GABA-evoked RDLbd I/V currents (pIC50 6.33), with no current from 10 uM ivermectin alone. The former positive-modulator EC50 contradicted that assay; an inhibitory model mapping has not yet been validated." ivermectin,vertebrate_GlyR,EC50,1e-06,1.0,positive_modulator,exact_compound_exact_receptor_other_species,E1,rat/human alpha1 glycine receptor,literature_order,functional_engagement_proxy,"Shan et al. 2001, J Biol Chem 276:12556 (ivermectin is an unconventional agonist of the alpha1 glycine receptor, sub-micromolar to micromolar)." ivermectin,vertebrate_GABA_A,EC50,1e-05,1.0,positive_modulator,exact_compound_exact_receptor_other_species,E1,Mus musculus GABA-A,literature_order,functional_engagement_proxy,"Dawson et al. 2000, J Pharmacol Exp Ther 295:1051 (avermectin effects in mice are mediated through GABA-A; micromolar potentiation)." ivermectin,insect_nAChR,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder -chlorpyrifos_oxon,insect_AChE,IC50,1e-08,1.0,inhibitor,exact_compound_exact_receptor_exact_species,E0,Drosophila melanogaster AChE (Ace),literature_order,functional_engagement,"Fournier and Mutero 1994, Comp Biochem Physiol C 108:19 (Drosophila AChE is inhibited by oxon organophosphates in the low-nM order and is more sensitive than mammalian AChE)." +chlorpyrifos_oxon,insect_AChE,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,"Placeholder pending source verification: Fournier and Mutero 1994 is a resistance review, not a primary Drosophila chlorpyrifos-oxon 10 nM IC50 assay. The previous number is withheld pending a primary enzyme assay with exposure time." chlorpyrifos_oxon,vertebrate_AChE,IC50,3e-07,1.0,inhibitor,exact_compound_exact_receptor_other_species,E1,Homo sapiens erythrocyte AChE in donor blood,literature_order,functional_engagement_proxy,"Archives of Toxicology 2020, interspecies AChE inhibition study in electric eel and human donor blood (chlorpyrifos-oxon IC50 about 0.3 uM in human blood)." chlorpyrifos_oxon,insect_nAChR,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,"class-order placeholder: the oxon acts on the enzyme, not directly on the receptor" chlorpyrifos_oxon,vertebrate_nAChR_a4b2,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder @@ -83,7 +83,7 @@ deltamethrin,insect_nAChR,unknown,,1.0,none,unsupported,E4,,class_placeholder,no deltamethrin,insect_RDL,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder deltamethrin,vertebrate_GABA_A,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder permethrin,insect_Nav,EC50,4e-07,1.0,positive_modulator,exact_compound_exact_receptor_exact_species,E0,Drosophila melanogaster para in Xenopus oocytes,literature_order,functional_engagement,"Burton et al. 2011, Insect Biochem Mol Biol 41:723 (Drosophila para tail-current EC50 0.40 uM for permethrin)." -permethrin,vertebrate_Nav1_x,EC50,4e-05,1.0,positive_modulator,class_extrapolation,E3,Rattus norvegicus brain type IIA sodium channels (fold-difference statement),literature_order,functional_engagement_proxy,"Smith and Soderlund 1998, Neurotoxicology 19:823 (permethrin is more than 100-fold less potent on rat brain type IIA sodium channels than on insect para)." +permethrin,vertebrate_Nav1_x,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,"Placeholder pending source verification: Smith and Soderlund 1998 (PMID 9863771) studied cypermethrin, not permethrin. It cannot support the former 40 uM permethrin parameter; a matched mammalian permethrin assay is required." permethrin,insect_nAChR,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder permethrin,insect_RDL,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder permethrin,vertebrate_GABA_A,unknown,,1.0,none,unsupported,E4,,class_placeholder,not_modelled,class-order placeholder diff --git a/papers/tables/T10_evidence_types.md b/papers/tables/T10_evidence_types.md index 7f662f4..b8e5c89 100644 --- a/papers/tables/T10_evidence_types.md +++ b/papers/tables/T10_evidence_types.md @@ -23,7 +23,7 @@ | fipronil | insect_nAChR | unknown | -- | E4 | not_modelled | | fipronil | vertebrate_nAChR_a4b2 | unknown | -- | E4 | not_modelled | | fipronil | vertebrate_GABA_A | IC50 | 1.10e-06 | E1 | functional_engagement_proxy | -| fipronil | insect_RDL | IC50 | 3.00e-08 | E0 | functional_engagement | +| fipronil | insect_RDL | IC50 | 3.00e-08 | E2 | functional_engagement_proxy | | clothianidin | insect_nAChR | EC50 | 3.00e-08 | E2 | functional_engagement_proxy | | clothianidin | vertebrate_nAChR_a7 | EC50 | 7.40e-04 | E1 | functional_engagement_proxy | | clothianidin | vertebrate_nAChR_a4b2 | EC50 | 1.00e-04 | E3 | functional_engagement_proxy | @@ -39,17 +39,17 @@ | acetamiprid | vertebrate_nAChR_a4b2 | EC50 | 1.00e-04 | E3 | functional_engagement_proxy | | acetamiprid | insect_RDL | unknown | -- | E4 | not_modelled | | acetamiprid | vertebrate_GABA_A | unknown | -- | E4 | not_modelled | -| sulfoxaflor | insect_nAChR | EC50 | 1.50e-04 | E2 | functional_engagement_proxy | +| sulfoxaflor | insect_nAChR | unknown | -- | E4 | not_modelled | | sulfoxaflor | vertebrate_nAChR_a4b2 | EC50 | 0.001 | E3 | functional_engagement_proxy | | sulfoxaflor | vertebrate_nAChR_a7 | unknown | -- | E4 | not_modelled | | sulfoxaflor | insect_RDL | unknown | -- | E4 | not_modelled | | sulfoxaflor | vertebrate_GABA_A | unknown | -- | E4 | not_modelled | -| spinosad | insect_nAChR | EC50 | 5.00e-06 | E2 | functional_engagement_proxy | +| spinosad | insect_nAChR | unknown | -- | E4 | not_modelled | | spinosad | vertebrate_nAChR_a4b2 | unknown | -- | E4 | not_modelled | | spinosad | vertebrate_nAChR_a7 | unknown | -- | E4 | not_modelled | | spinosad | insect_RDL | unknown | -- | E4 | not_modelled | | spinosad | vertebrate_GABA_A | unknown | -- | E4 | not_modelled | -| spinosad | insect_nAChR_alpha6 | EC50 | 5.00e-06 | E2 | functional_engagement_proxy | +| spinosad | insect_nAChR_alpha6 | unknown | -- | E4 | not_modelled | | acetylcholine | insect_nAChR | EC50 | 1.00e-05 | E2 | functional_engagement_proxy | | acetylcholine | vertebrate_nAChR_a4b2 | EC50 | 1.00e-05 | E1 | functional_engagement_proxy | | acetylcholine | vertebrate_nAChR_a7 | EC50 | 1.60e-04 | E1 | functional_engagement_proxy | @@ -62,7 +62,7 @@ | gaba | vertebrate_nAChR_a4b2 | unknown | -- | E4 | not_modelled | | picrotoxin | insect_RDL | IC50 | 1.00e-06 | E0 | functional_engagement | | picrotoxin | vertebrate_GABA_A | IC50 | 1.00e-06 | E1 | functional_engagement_proxy | -| picrotoxin | vertebrate_GlyR | IC50 | 1.00e-05 | E3 | functional_engagement_proxy | +| picrotoxin | vertebrate_GlyR | unknown | -- | E4 | not_modelled | | picrotoxin | insect_nAChR | unknown | -- | E4 | not_modelled | | picrotoxin | vertebrate_nAChR_a4b2 | unknown | -- | E4 | not_modelled | | dieldrin | insect_RDL | IC50 | 1.00e-07 | E0 | functional_engagement | @@ -70,11 +70,11 @@ | dieldrin | insect_nAChR | unknown | -- | E4 | not_modelled | | dieldrin | vertebrate_nAChR_a4b2 | unknown | -- | E4 | not_modelled | | ivermectin | insect_GluCl | EC50 | 2.00e-08 | E2 | functional_engagement_proxy | -| ivermectin | insect_RDL | EC50 | 1.00e-06 | E0 | functional_engagement | +| ivermectin | insect_RDL | unknown | -- | E4 | not_modelled | | ivermectin | vertebrate_GlyR | EC50 | 1.00e-06 | E1 | functional_engagement_proxy | | ivermectin | vertebrate_GABA_A | EC50 | 1.00e-05 | E1 | functional_engagement_proxy | | ivermectin | insect_nAChR | unknown | -- | E4 | not_modelled | -| chlorpyrifos_oxon | insect_AChE | IC50 | 1.00e-08 | E0 | functional_engagement | +| chlorpyrifos_oxon | insect_AChE | unknown | -- | E4 | not_modelled | | chlorpyrifos_oxon | vertebrate_AChE | IC50 | 3.00e-07 | E1 | functional_engagement_proxy | | chlorpyrifos_oxon | insect_nAChR | unknown | -- | E4 | not_modelled | | chlorpyrifos_oxon | vertebrate_nAChR_a4b2 | unknown | -- | E4 | not_modelled | @@ -86,7 +86,7 @@ | deltamethrin | insect_RDL | unknown | -- | E4 | not_modelled | | deltamethrin | vertebrate_GABA_A | unknown | -- | E4 | not_modelled | | permethrin | insect_Nav | EC50 | 4.00e-07 | E0 | functional_engagement | -| permethrin | vertebrate_Nav1_x | EC50 | 4.00e-05 | E3 | functional_engagement_proxy | +| permethrin | vertebrate_Nav1_x | unknown | -- | E4 | not_modelled | | permethrin | insect_nAChR | unknown | -- | E4 | not_modelled | | permethrin | insect_RDL | unknown | -- | E4 | not_modelled | | permethrin | vertebrate_GABA_A | unknown | -- | E4 | not_modelled | diff --git a/papers/tables/T12_dependence_landscape.csv b/papers/tables/T12_dependence_landscape.csv index 8731337..d34f191 100644 --- a/papers/tables/T12_dependence_landscape.csv +++ b/papers/tables/T12_dependence_landscape.csv @@ -1,12 +1,12 @@ compound,conc_M,real_effect,vehicle,effect_floor,delta,class,class_raw,class_absolute_floor,q_value,necessary_level,necessary_level_verdict,non_monotone,p_sign,p_sign_wm,p_weight,p_degree,p_ER,q_weight,q_degree,verdict_weight,verdict_degree -acetamiprid,1e-08,0.19091781755964465,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5448090371167295,degree_sequence,equivalent_within_tolerance,False,0.47352647352647353,0.2837162837162837,0.2007992007992008,0.945054945054945,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -acetamiprid,1e-07,-3.377888997795963,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.32167832167832167,0.3196803196803197,0.3926073926073926,0.8471528471528471,0.000999000999000999,0.9038301148392892,1.0,equivalent_within_tolerance,equivalent_within_tolerance -acetamiprid,1e-06,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -acetamiprid,1e-05,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -acetylcholine,1e-08,0.0007607670748654982,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.5448090371167295,none_required,,False,0.44655344655344653,0.2837162837162837,0.2057942057942058,0.952047952047952,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -acetylcholine,1e-07,0.011839023579403474,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.5448090371167295,none_required,,False,0.44955044955044954,0.2837162837162837,0.2057942057942058,0.952047952047952,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -acetylcholine,1e-06,0.13799221453647537,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5448090371167295,degree_sequence,equivalent_within_tolerance,False,0.4695304695304695,0.2837162837162837,0.2017982017982018,0.945054945054945,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -acetylcholine,1e-05,-1.4378006330560869,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.6636999364272091,degree_sequence,equivalent_within_tolerance,False,0.35064935064935066,0.2937062937062937,0.26073926073926074,0.978021978021978,0.000999000999000999,0.6636999364272091,1.0,equivalent_within_tolerance,equivalent_within_tolerance +acetamiprid,1e-08,0.19091781755964465,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.6557886557886557,degree_sequence,equivalent_within_tolerance,False,0.47352647352647353,0.2837162837162837,0.2007992007992008,0.945054945054945,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance +acetamiprid,1e-07,-3.377888997795963,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.32167832167832167,0.3196803196803197,0.3926073926073926,0.8471528471528471,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +acetamiprid,1e-06,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +acetamiprid,1e-05,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +acetylcholine,1e-08,0.0007607670748654982,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.6557886557886557,none_required,,False,0.44655344655344653,0.2837162837162837,0.2057942057942058,0.952047952047952,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance +acetylcholine,1e-07,0.011839023579403474,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.6557886557886557,none_required,,False,0.44955044955044954,0.2837162837162837,0.2057942057942058,0.952047952047952,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance +acetylcholine,1e-06,0.13799221453647537,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.6557886557886557,degree_sequence,equivalent_within_tolerance,False,0.4695304695304695,0.2837162837162837,0.2017982017982018,0.945054945054945,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance +acetylcholine,1e-05,-1.4378006330560869,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.796439923712651,degree_sequence,equivalent_within_tolerance,False,0.35064935064935066,0.2937062937062937,0.26073926073926074,0.978021978021978,0.000999000999000999,0.796439923712651,1.0,equivalent_within_tolerance,equivalent_within_tolerance caffeine,1e-08,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance caffeine,1e-07,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance caffeine,1e-06,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance @@ -15,71 +15,71 @@ chlordimeform,1e-08,0.0,6.631244285926712,0.06631244285926713,0.3315622142963356 chlordimeform,1e-07,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance chlordimeform,1e-06,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance chlordimeform,1e-05,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance -chlorpyrifos_oxon,1e-08,41.89007154138836,6.631244285926712,0.06631244285926713,0.33156221429633564,topology-dependent,topology-dependent,topology-dependent,0.04936240230357877,degree_sequence,indeterminate,True,0.10989010989010989,0.2807192807192807,0.000999000999000999,0.04295704295704296,0.000999000999000999,0.04936240230357877,0.21869040050858235,distinguishable,indeterminate -chlorpyrifos_oxon,1e-07,-4.453314692141862,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.3026973026973027,0.33666333666333664,0.4855144855144855,0.7302697302697303,0.000999000999000999,0.9038301148392892,1.0,equivalent_within_tolerance,equivalent_within_tolerance -chlorpyrifos_oxon,1e-06,-5.555436983853854,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.2887112887112887,0.37162837162837165,0.5644355644355644,0.5604395604395604,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -chlorpyrifos_oxon,1e-05,-5.549928143951719,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.2887112887112887,0.3706293706293706,0.5644355644355644,0.5604395604395604,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -clothianidin,1e-08,0.10000529217663079,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5448090371167295,degree_sequence,equivalent_within_tolerance,False,0.46153846153846156,0.2837162837162837,0.20279720279720279,0.9460539460539461,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -clothianidin,1e-07,-4.773623977900712,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.2997002997002997,0.34165834165834164,0.5084915084915085,0.6903096903096904,0.000999000999000999,0.9038301148392892,1.0,equivalent_within_tolerance,equivalent_within_tolerance -clothianidin,1e-06,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -clothianidin,1e-05,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -ddt,1e-08,0.003091907228121471,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.9038301148392892,none_required,,False,0.3596403596403596,0.28771228771228774,0.5514485514485514,0.962037962037962,0.000999000999000999,0.9038301148392892,1.0,equivalent_within_tolerance,equivalent_within_tolerance -ddt,1e-07,0.030911376259251888,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.9038301148392892,none_required,,False,0.3596403596403596,0.2867132867132867,0.5514485514485514,0.9600399600399601,0.000999000999000999,0.9038301148392892,1.0,equivalent_within_tolerance,equivalent_within_tolerance -ddt,1e-06,0.30831405448478044,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.3646353646353646,0.2857142857142857,0.5514485514485514,0.9470529470529471,0.000999000999000999,0.9038301148392892,1.0,equivalent_within_tolerance,equivalent_within_tolerance -ddt,1e-05,2.9816464493304498,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.5044955044955045,0.27372627372627373,0.5514485514485514,0.8691308691308691,0.000999000999000999,0.9038301148392892,1.0,equivalent_within_tolerance,equivalent_within_tolerance -deltamethrin,1e-08,4.40482935703646,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.5924075924075924,0.26373626373626374,0.5434565434565435,0.8431568431568431,0.000999000999000999,0.9038301148392892,1.0,equivalent_within_tolerance,equivalent_within_tolerance -deltamethrin,1e-07,24.021540644552797,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5072261072261073,degree_sequence,equivalent_within_tolerance,False,0.5104895104895105,0.14385614385614387,0.13586413586413587,0.936063936063936,0.000999000999000999,0.5072261072261073,1.0,indeterminate,equivalent_within_tolerance -deltamethrin,1e-06,37.15887655509541,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.32727272727272727,degree_sequence,indeterminate,False,0.22977022977022976,0.3166833166833167,0.07692307692307693,0.43756243756243757,0.000999000999000999,0.32727272727272727,0.9038301148392892,indeterminate,indeterminate -deltamethrin,1e-05,38.16064501855905,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.39560439560439553,degree_sequence,indeterminate,False,0.21978021978021978,0.6193806193806194,0.0989010989010989,0.37462537462537465,0.000999000999000999,0.39560439560439553,0.9038301148392892,indeterminate,indeterminate +chlorpyrifos_oxon,1e-08,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +chlorpyrifos_oxon,1e-07,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +chlorpyrifos_oxon,1e-06,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +chlorpyrifos_oxon,1e-05,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +clothianidin,1e-08,0.10000529217663079,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.6557886557886557,degree_sequence,equivalent_within_tolerance,False,0.46153846153846156,0.2837162837162837,0.20279720279720279,0.9460539460539461,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance +clothianidin,1e-07,-4.773623977900712,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.2997002997002997,0.34165834165834164,0.5084915084915085,0.6903096903096904,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +clothianidin,1e-06,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +clothianidin,1e-05,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +ddt,1e-08,0.003091907228121471,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,1.0,none_required,,False,0.3596403596403596,0.28771228771228774,0.5514485514485514,0.962037962037962,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +ddt,1e-07,0.030911376259251888,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,1.0,none_required,,False,0.3596403596403596,0.2867132867132867,0.5514485514485514,0.9600399600399601,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +ddt,1e-06,0.30831405448478044,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.3646353646353646,0.2857142857142857,0.5514485514485514,0.9470529470529471,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +ddt,1e-05,2.9816464493304498,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.5044955044955045,0.27372627372627373,0.5514485514485514,0.8691308691308691,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +deltamethrin,1e-08,4.40482935703646,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.5924075924075924,0.26373626373626374,0.5434565434565435,0.8431568431568431,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +deltamethrin,1e-07,24.021540644552797,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5567115811018251,degree_sequence,equivalent_within_tolerance,False,0.5104895104895105,0.14385614385614387,0.13586413586413587,0.936063936063936,0.000999000999000999,0.5567115811018251,1.0,indeterminate,equivalent_within_tolerance +deltamethrin,1e-06,37.15887655509541,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.34449760765550236,degree_sequence,indeterminate,False,0.22977022977022976,0.3166833166833167,0.07692307692307693,0.43756243756243757,0.000999000999000999,0.34449760765550236,1.0,indeterminate,indeterminate +deltamethrin,1e-05,38.16064501855905,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.42603550295857984,degree_sequence,indeterminate,False,0.21978021978021978,0.6193806193806194,0.0989010989010989,0.37462537462537465,0.000999000999000999,0.42603550295857984,1.0,indeterminate,indeterminate diazepam,1e-08,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance diazepam,1e-07,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance diazepam,1e-06,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance diazepam,1e-05,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance -dieldrin,1e-08,0.07085762022170528,6.631244285926712,0.06631244285926713,0.33156221429633564,topology-dependent,topology-dependent,topology-dependent,0.04936240230357877,degree_sequence,equivalent_within_tolerance,True,0.07792207792207792,0.47152847152847155,0.004995004995004995,0.04295704295704296,0.000999000999000999,0.04936240230357877,0.21869040050858235,distinguishable,equivalent_within_tolerance -dieldrin,1e-07,0.4236252131450913,6.631244285926712,0.06631244285926713,0.33156221429633564,topology-dependent,topology-dependent,topology-dependent,0.04936240230357877,degree_sequence,equivalent_within_tolerance,True,0.0959040959040959,0.3386613386613387,0.004995004995004995,0.023976023976023976,0.000999000999000999,0.04936240230357877,0.13889558717144926,distinguishable,equivalent_within_tolerance -dieldrin,1e-06,0.8512734176258983,6.631244285926712,0.06631244285926713,0.33156221429633564,topology-dependent,topology-dependent,topology-dependent,0.04936240230357877,degree_sequence,equivalent_within_tolerance,True,0.11388611388611389,0.2177822177822178,0.004995004995004995,0.007992007992007992,0.000999000999000999,0.04936240230357877,0.053706293706293706,distinguishable,equivalent_within_tolerance -dieldrin,1e-05,0.8996076235547861,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.05034965034965035,degree_sequence,equivalent_within_tolerance,False,0.11288711288711288,0.2097902097902098,0.005994005994005994,0.006993006993006993,0.000999000999000999,0.05034965034965035,0.05107935542718151,equivalent_within_tolerance,equivalent_within_tolerance -fipronil,1e-08,0.16838507486121923,6.631244285926712,0.06631244285926713,0.33156221429633564,topology-dependent,topology-dependent,topology-dependent,0.04936240230357877,degree_sequence,equivalent_within_tolerance,True,0.08391608391608392,0.4405594405594406,0.004995004995004995,0.03796203796203796,0.000999000999000999,0.04936240230357877,0.2125874125874126,distinguishable,equivalent_within_tolerance -fipronil,1e-07,0.7380224886401772,6.631244285926712,0.06631244285926713,0.33156221429633564,topology-dependent,topology-dependent,topology-dependent,0.04936240230357877,degree_sequence,equivalent_within_tolerance,True,0.11288711288711288,0.24375624375624375,0.004995004995004995,0.01098901098901099,0.000999000999000999,0.04936240230357877,0.06837606837606838,distinguishable,equivalent_within_tolerance -fipronil,1e-06,0.8996076235547861,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.05034965034965035,degree_sequence,equivalent_within_tolerance,False,0.11288711288711288,0.2097902097902098,0.005994005994005994,0.006993006993006993,0.000999000999000999,0.05034965034965035,0.05107935542718151,equivalent_within_tolerance,equivalent_within_tolerance -fipronil,1e-05,0.8996076235547861,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.05034965034965035,degree_sequence,equivalent_within_tolerance,False,0.11288711288711288,0.2097902097902098,0.005994005994005994,0.006993006993006993,0.000999000999000999,0.05034965034965035,0.05107935542718151,equivalent_within_tolerance,equivalent_within_tolerance -gaba,1e-08,-8.671055358178137e-07,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,0.04936240230357877,none_required,,False,0.07692307692307693,0.4955044955044955,0.004995004995004995,0.05194805194805195,0.000999000999000999,0.04936240230357877,0.22966507177033496,distinguishable,equivalent_within_tolerance -gaba,1e-07,-2.741773682846116e-05,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,topology-dependent,0.04936240230357877,none_required,,False,0.07692307692307693,0.4955044955044955,0.004995004995004995,0.05194805194805195,0.000999000999000999,0.04936240230357877,0.22966507177033496,distinguishable,equivalent_within_tolerance -gaba,1e-06,-0.0008644848379635306,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,topology-dependent,0.04936240230357877,none_required,,False,0.07692307692307693,0.4955044955044955,0.004995004995004995,0.05194805194805195,0.000999000999000999,0.04936240230357877,0.22966507177033496,distinguishable,equivalent_within_tolerance -gaba,1e-05,-0.025019861031331914,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,topology-dependent,0.04936240230357877,none_required,,False,0.07692307692307693,0.5054945054945055,0.004995004995004995,0.05194805194805195,0.000999000999000999,0.04936240230357877,0.22966507177033496,distinguishable,equivalent_within_tolerance -imidacloprid,1e-08,-0.19448112604907042,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5448090371167295,degree_sequence,equivalent_within_tolerance,False,0.42357642357642356,0.2857142857142857,0.21078921078921078,0.958041958041958,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -imidacloprid,1e-07,-5.680388639120631,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.2857142857142857,0.3776223776223776,0.5674325674325674,0.5374625374625375,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -imidacloprid,1e-06,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -imidacloprid,1e-05,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -ivermectin,1e-08,-0.015564152084219351,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.064550833781603,none_required,,False,0.8151848151848152,0.7882117882117882,0.00999000999000999,0.5634365634365635,0.000999000999000999,0.064550833781603,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -ivermectin,1e-07,-0.0587293185688571,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,topology-dependent,0.04936240230357877,none_required,,False,0.5344655344655345,0.9240759240759241,0.003996003996003996,0.2957042957042957,0.000999000999000999,0.04936240230357877,0.7414674877361443,distinguishable,equivalent_within_tolerance -ivermectin,1e-06,-0.1835528137105662,6.631244285926712,0.06631244285926713,0.33156221429633564,topology-dependent,topology-dependent,topology-dependent,0.04936240230357877,degree_sequence,equivalent_within_tolerance,True,0.2047952047952048,0.6193806193806194,0.001998001998001998,0.1028971028971029,0.000999000999000999,0.04936240230357877,0.4020165880630997,distinguishable,equivalent_within_tolerance -ivermectin,1e-05,-0.29716171988116447,6.631244285926712,0.06631244285926713,0.33156221429633564,topology-dependent,topology-dependent,topology-dependent,0.04936240230357877,degree_sequence,equivalent_within_tolerance,True,0.12287712287712288,0.6023976023976024,0.001998001998001998,0.08591408591408592,0.000999000999000999,0.04936240230357877,0.3520382056967423,distinguishable,equivalent_within_tolerance -nicotine,1e-08,0.09320704732198237,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5448090371167295,degree_sequence,equivalent_within_tolerance,False,0.46053946053946054,0.2837162837162837,0.20279720279720279,0.9470529470529471,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -nicotine,1e-07,-0.26018164807203803,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5448090371167295,degree_sequence,equivalent_within_tolerance,False,0.4165834165834166,0.2857142857142857,0.21078921078921078,0.9600399600399601,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -nicotine,1e-06,-5.40399067063418,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.2887112887112887,0.3626373626373626,0.5494505494505495,0.5794205794205795,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -nicotine,1e-05,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -nitenpyram,1e-08,0.19091781755964465,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5448090371167295,degree_sequence,equivalent_within_tolerance,False,0.47352647352647353,0.2837162837162837,0.2007992007992008,0.945054945054945,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -nitenpyram,1e-07,-3.377888997795963,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.32167832167832167,0.3196803196803197,0.3926073926073926,0.8471528471528471,0.000999000999000999,0.9038301148392892,1.0,equivalent_within_tolerance,equivalent_within_tolerance -nitenpyram,1e-06,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -nitenpyram,1e-05,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,0.9038301148392892,0.9038301148392892,equivalent_within_tolerance,equivalent_within_tolerance -permethrin,1e-08,0.5000584660495315,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.3696303696303696,0.2857142857142857,0.5534465534465535,0.936063936063936,0.000999000999000999,0.9038301148392892,1.0,equivalent_within_tolerance,equivalent_within_tolerance -permethrin,1e-07,4.710059982330625,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.9038301148392892,degree_sequence,equivalent_within_tolerance,False,0.6153846153846154,0.2627372627372627,0.5404595404595405,0.8351648351648352,0.000999000999000999,0.9038301148392892,1.0,equivalent_within_tolerance,equivalent_within_tolerance -permethrin,1e-06,24.820376182767554,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.4615384615384615,degree_sequence,equivalent_within_tolerance,False,0.4985014985014985,0.14085914085914086,0.12087912087912088,0.961038961038961,0.000999000999000999,0.4615384615384615,1.0,indeterminate,equivalent_within_tolerance -permethrin,1e-05,37.24852214329439,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.32727272727272727,degree_sequence,indeterminate,False,0.22877122877122877,0.34665334665334663,0.07792207792207792,0.43156843156843155,0.000999000999000999,0.32727272727272727,0.9038301148392892,indeterminate,indeterminate -picrotoxin,1e-08,0.00760206859210566,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,topology-dependent,0.04936240230357877,none_required,,False,0.07692307692307693,0.4935064935064935,0.004995004995004995,0.05094905094905095,0.000999000999000999,0.04936240230357877,0.22966507177033496,distinguishable,equivalent_within_tolerance -picrotoxin,1e-07,0.07085762022170528,6.631244285926712,0.06631244285926713,0.33156221429633564,topology-dependent,topology-dependent,topology-dependent,0.04936240230357877,degree_sequence,equivalent_within_tolerance,True,0.07792207792207792,0.47152847152847155,0.004995004995004995,0.04295704295704296,0.000999000999000999,0.04936240230357877,0.21869040050858235,distinguishable,equivalent_within_tolerance -picrotoxin,1e-06,0.4236252131450913,6.631244285926712,0.06631244285926713,0.33156221429633564,topology-dependent,topology-dependent,topology-dependent,0.04936240230357877,degree_sequence,equivalent_within_tolerance,True,0.0959040959040959,0.3386613386613387,0.004995004995004995,0.023976023976023976,0.000999000999000999,0.04936240230357877,0.13889558717144926,distinguishable,equivalent_within_tolerance -picrotoxin,1e-05,0.8512734176258983,6.631244285926712,0.06631244285926713,0.33156221429633564,topology-dependent,topology-dependent,topology-dependent,0.04936240230357877,degree_sequence,equivalent_within_tolerance,True,0.11388611388611389,0.2177822177822178,0.004995004995004995,0.007992007992007992,0.000999000999000999,0.04936240230357877,0.053706293706293706,distinguishable,equivalent_within_tolerance -spinosad,1e-08,0.006005747612804591,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.5448090371167295,none_required,,False,0.4485514485514486,0.2837162837162837,0.2057942057942058,0.952047952047952,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -spinosad,1e-07,0.054912834542182765,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.5448090371167295,none_required,,False,0.45454545454545453,0.2837162837162837,0.2047952047952048,0.9500499500499501,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -spinosad,1e-06,0.16958996890116396,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5448090371167295,degree_sequence,equivalent_within_tolerance,False,0.4725274725274725,0.2837162837162837,0.2017982017982018,0.945054945054945,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -spinosad,1e-05,-3.0409209448354613,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.8909913615795969,degree_sequence,equivalent_within_tolerance,False,0.32367632367632365,0.31368631368631367,0.36063936063936064,0.8701298701298701,0.000999000999000999,0.8909913615795969,1.0,equivalent_within_tolerance,equivalent_within_tolerance -sulfoxaflor,1e-08,0.00020209322619546555,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.5448090371167295,none_required,,False,0.44655344655344653,0.2837162837162837,0.2057942057942058,0.952047952047952,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -sulfoxaflor,1e-07,0.0020150179661975187,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.5448090371167295,none_required,,False,0.44755244755244755,0.2837162837162837,0.2057942057942058,0.952047952047952,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -sulfoxaflor,1e-06,0.019564984263337415,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.5448090371167295,none_required,,False,0.45054945054945056,0.2837162837162837,0.2057942057942058,0.952047952047952,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -sulfoxaflor,1e-05,0.14294144232433847,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5448090371167295,degree_sequence,equivalent_within_tolerance,False,0.47052947052947053,0.2837162837162837,0.2017982017982018,0.945054945054945,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -thiamethoxam,1e-08,0.00030309050857635356,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.5448090371167295,none_required,,False,0.44655344655344653,0.2837162837162837,0.2057942057942058,0.952047952047952,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -thiamethoxam,1e-07,0.0030176057607382845,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.5448090371167295,none_required,,False,0.44755244755244755,0.2837162837162837,0.2057942057942058,0.952047952047952,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -thiamethoxam,1e-06,0.028867082651210296,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.5448090371167295,none_required,,False,0.45254745254745254,0.2837162837162837,0.2057942057942058,0.951048951048951,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance -thiamethoxam,1e-05,0.1766469540699287,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5448090371167295,degree_sequence,equivalent_within_tolerance,False,0.4725274725274725,0.2837162837162837,0.2017982017982018,0.945054945054945,0.000999000999000999,0.5448090371167295,1.0,equivalent_within_tolerance,equivalent_within_tolerance +dieldrin,1e-08,0.07085762022170528,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.061832903938167094,degree_sequence,equivalent_within_tolerance,False,0.07792207792207792,0.47152847152847155,0.004995004995004995,0.04295704295704296,0.000999000999000999,0.061832903938167094,0.23279945860591025,equivalent_within_tolerance,equivalent_within_tolerance +dieldrin,1e-07,0.4236252131450913,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.061832903938167094,degree_sequence,equivalent_within_tolerance,False,0.0959040959040959,0.3386613386613387,0.004995004995004995,0.023976023976023976,0.000999000999000999,0.061832903938167094,0.14385614385614387,equivalent_within_tolerance,equivalent_within_tolerance +dieldrin,1e-06,0.8512734176258983,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.061832903938167094,degree_sequence,equivalent_within_tolerance,False,0.11388611388611389,0.2177822177822178,0.004995004995004995,0.007992007992007992,0.000999000999000999,0.061832903938167094,0.06393606393606394,equivalent_within_tolerance,equivalent_within_tolerance +dieldrin,1e-05,0.8996076235547861,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.061832903938167094,degree_sequence,equivalent_within_tolerance,False,0.11288711288711288,0.2097902097902098,0.005994005994005994,0.006993006993006993,0.000999000999000999,0.061832903938167094,0.061832903938167094,equivalent_within_tolerance,equivalent_within_tolerance +fipronil,1e-08,0.16838507486121923,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.061832903938167094,degree_sequence,equivalent_within_tolerance,False,0.08391608391608392,0.4405594405594406,0.004995004995004995,0.03796203796203796,0.000999000999000999,0.061832903938167094,0.2199180130214613,equivalent_within_tolerance,equivalent_within_tolerance +fipronil,1e-07,0.7380224886401772,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.061832903938167094,degree_sequence,equivalent_within_tolerance,False,0.11288711288711288,0.24375624375624375,0.004995004995004995,0.01098901098901099,0.000999000999000999,0.061832903938167094,0.07100591715976332,equivalent_within_tolerance,equivalent_within_tolerance +fipronil,1e-06,0.8996076235547861,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.061832903938167094,degree_sequence,equivalent_within_tolerance,False,0.11288711288711288,0.2097902097902098,0.005994005994005994,0.006993006993006993,0.000999000999000999,0.061832903938167094,0.061832903938167094,equivalent_within_tolerance,equivalent_within_tolerance +fipronil,1e-05,0.8996076235547861,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.061832903938167094,degree_sequence,equivalent_within_tolerance,False,0.11288711288711288,0.2097902097902098,0.005994005994005994,0.006993006993006993,0.000999000999000999,0.061832903938167094,0.061832903938167094,equivalent_within_tolerance,equivalent_within_tolerance +gaba,1e-08,-8.671055358178137e-07,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,0.061832903938167094,none_required,,False,0.07692307692307693,0.4955044955044955,0.004995004995004995,0.05194805194805195,0.000999000999000999,0.061832903938167094,0.24242424242424246,equivalent_within_tolerance,equivalent_within_tolerance +gaba,1e-07,-2.741773682846116e-05,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.061832903938167094,none_required,,False,0.07692307692307693,0.4955044955044955,0.004995004995004995,0.05194805194805195,0.000999000999000999,0.061832903938167094,0.24242424242424246,equivalent_within_tolerance,equivalent_within_tolerance +gaba,1e-06,-0.0008644848379635306,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.061832903938167094,none_required,,False,0.07692307692307693,0.4955044955044955,0.004995004995004995,0.05194805194805195,0.000999000999000999,0.061832903938167094,0.24242424242424246,equivalent_within_tolerance,equivalent_within_tolerance +gaba,1e-05,-0.025019861031331914,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.061832903938167094,none_required,,False,0.07692307692307693,0.5054945054945055,0.004995004995004995,0.05194805194805195,0.000999000999000999,0.061832903938167094,0.24242424242424246,equivalent_within_tolerance,equivalent_within_tolerance +imidacloprid,1e-08,-0.19448112604907042,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.6557886557886557,degree_sequence,equivalent_within_tolerance,False,0.42357642357642356,0.2857142857142857,0.21078921078921078,0.958041958041958,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance +imidacloprid,1e-07,-5.680388639120631,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.2857142857142857,0.3776223776223776,0.5674325674325674,0.5374625374625375,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +imidacloprid,1e-06,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +imidacloprid,1e-05,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +ivermectin,1e-08,-0.012538678824231653,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.07100591715976332,none_required,,False,0.906093906093906,0.7042957042957043,0.01098901098901099,0.8331668331668332,0.000999000999000999,0.07100591715976332,1.0,equivalent_within_tolerance,equivalent_within_tolerance +ivermectin,1e-07,-0.03121681365661999,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.07100591715976332,none_required,,False,0.8811188811188811,0.7072927072927073,0.01098901098901099,0.8391608391608392,0.000999000999000999,0.07100591715976332,1.0,equivalent_within_tolerance,equivalent_within_tolerance +ivermectin,1e-06,-0.03668098746430726,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.07100591715976332,none_required,,False,0.8731268731268731,0.7102897102897103,0.01098901098901099,0.8411588411588412,0.000999000999000999,0.07100591715976332,1.0,equivalent_within_tolerance,equivalent_within_tolerance +ivermectin,1e-05,-0.03733448908736925,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.07100591715976332,none_required,,False,0.8721278721278721,0.7102897102897103,0.01098901098901099,0.8411588411588412,0.000999000999000999,0.07100591715976332,1.0,equivalent_within_tolerance,equivalent_within_tolerance +nicotine,1e-08,0.09320704732198237,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.6557886557886557,degree_sequence,equivalent_within_tolerance,False,0.46053946053946054,0.2837162837162837,0.20279720279720279,0.9470529470529471,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance +nicotine,1e-07,-0.26018164807203803,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.6557886557886557,degree_sequence,equivalent_within_tolerance,False,0.4165834165834166,0.2857142857142857,0.21078921078921078,0.9600399600399601,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance +nicotine,1e-06,-5.40399067063418,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.2887112887112887,0.3626373626373626,0.5494505494505495,0.5794205794205795,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +nicotine,1e-05,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +nitenpyram,1e-08,0.19091781755964465,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.6557886557886557,degree_sequence,equivalent_within_tolerance,False,0.47352647352647353,0.2837162837162837,0.2007992007992008,0.945054945054945,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance +nitenpyram,1e-07,-3.377888997795963,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.32167832167832167,0.3196803196803197,0.3926073926073926,0.8471528471528471,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +nitenpyram,1e-06,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +nitenpyram,1e-05,-6.17278216613422,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.27472527472527475,0.3886113886113886,0.5864135864135864,0.47152847152847155,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +permethrin,1e-08,0.5000584660495315,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.3696303696303696,0.2857142857142857,0.5534465534465535,0.936063936063936,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +permethrin,1e-07,4.710059982330625,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,1.0,degree_sequence,equivalent_within_tolerance,False,0.6153846153846154,0.2627372627372627,0.5404595404595405,0.8351648351648352,0.000999000999000999,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +permethrin,1e-06,24.820376182767554,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.5076923076923077,degree_sequence,equivalent_within_tolerance,False,0.4985014985014985,0.14085914085914086,0.12087912087912088,0.961038961038961,0.000999000999000999,0.5076923076923077,1.0,indeterminate,equivalent_within_tolerance +permethrin,1e-05,37.24852214329439,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.34449760765550236,degree_sequence,indeterminate,False,0.22877122877122877,0.34665334665334663,0.07792207792207792,0.43156843156843155,0.000999000999000999,0.34449760765550236,1.0,indeterminate,indeterminate +picrotoxin,1e-08,0.00760206859210566,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.061832903938167094,none_required,,False,0.07692307692307693,0.4935064935064935,0.004995004995004995,0.05094905094905095,0.000999000999000999,0.061832903938167094,0.24242424242424246,equivalent_within_tolerance,equivalent_within_tolerance +picrotoxin,1e-07,0.07085762022170528,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.061832903938167094,degree_sequence,equivalent_within_tolerance,False,0.07792207792207792,0.47152847152847155,0.004995004995004995,0.04295704295704296,0.000999000999000999,0.061832903938167094,0.23279945860591025,equivalent_within_tolerance,equivalent_within_tolerance +picrotoxin,1e-06,0.4236252131450913,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.061832903938167094,degree_sequence,equivalent_within_tolerance,False,0.0959040959040959,0.3386613386613387,0.004995004995004995,0.023976023976023976,0.000999000999000999,0.061832903938167094,0.14385614385614387,equivalent_within_tolerance,equivalent_within_tolerance +picrotoxin,1e-05,0.8512734176258983,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,topology-dependent,composition-dominated,0.061832903938167094,degree_sequence,equivalent_within_tolerance,False,0.11388611388611389,0.2177822177822178,0.004995004995004995,0.007992007992007992,0.000999000999000999,0.061832903938167094,0.06393606393606394,equivalent_within_tolerance,equivalent_within_tolerance +spinosad,1e-08,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +spinosad,1e-07,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +spinosad,1e-06,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +spinosad,1e-05,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +sulfoxaflor,1e-08,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +sulfoxaflor,1e-07,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +sulfoxaflor,1e-06,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +sulfoxaflor,1e-05,0.0,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +thiamethoxam,1e-08,0.00030309050857635356,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.6557886557886557,none_required,,False,0.44655344655344653,0.2837162837162837,0.2057942057942058,0.952047952047952,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance +thiamethoxam,1e-07,0.0030176057607382845,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.6557886557886557,none_required,,False,0.44755244755244755,0.2837162837162837,0.2057942057942058,0.952047952047952,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance +thiamethoxam,1e-06,0.028867082651210296,6.631244285926712,0.06631244285926713,0.33156221429633564,no-effect,no-effect,composition-dominated,0.6557886557886557,none_required,,False,0.45254745254745254,0.2837162837162837,0.2057942057942058,0.951048951048951,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance +thiamethoxam,1e-05,0.1766469540699287,6.631244285926712,0.06631244285926713,0.33156221429633564,composition-dominated,composition-dominated,composition-dominated,0.6557886557886557,degree_sequence,equivalent_within_tolerance,False,0.4725274725274725,0.2837162837162837,0.2017982017982018,0.945054945054945,0.000999000999000999,0.6557886557886557,1.0,equivalent_within_tolerance,equivalent_within_tolerance diff --git a/papers/tables/T12_dependence_landscape.md b/papers/tables/T12_dependence_landscape.md index dc7938c..8e0c8c9 100644 --- a/papers/tables/T12_dependence_landscape.md +++ b/papers/tables/T12_dependence_landscape.md @@ -2,14 +2,14 @@ | compound | conc_M | real_effect | class | class_raw | q_value | necessary_level | non_monotone | |---|---|---|---|---|---|---|---| -| acetamiprid | 1.00e-08 | 0.191 | composition-dominated | composition-dominated | 0.545 | degree_sequence | False | -| acetamiprid | 1.00e-07 | -3.378 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| acetamiprid | 1.00e-06 | -6.173 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| acetamiprid | 1.00e-05 | -6.173 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| acetylcholine | 1.00e-08 | 7.61e-04 | no-effect | no-effect | 0.545 | none_required | False | -| acetylcholine | 1.00e-07 | 0.012 | no-effect | no-effect | 0.545 | none_required | False | -| acetylcholine | 1.00e-06 | 0.138 | composition-dominated | composition-dominated | 0.545 | degree_sequence | False | -| acetylcholine | 1.00e-05 | -1.438 | composition-dominated | composition-dominated | 0.664 | degree_sequence | False | +| acetamiprid | 1.00e-08 | 0.191 | composition-dominated | composition-dominated | 0.656 | degree_sequence | False | +| acetamiprid | 1.00e-07 | -3.378 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| acetamiprid | 1.00e-06 | -6.173 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| acetamiprid | 1.00e-05 | -6.173 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| acetylcholine | 1.00e-08 | 7.61e-04 | no-effect | no-effect | 0.656 | none_required | False | +| acetylcholine | 1.00e-07 | 0.012 | no-effect | no-effect | 0.656 | none_required | False | +| acetylcholine | 1.00e-06 | 0.138 | composition-dominated | composition-dominated | 0.656 | degree_sequence | False | +| acetylcholine | 1.00e-05 | -1.438 | composition-dominated | composition-dominated | 0.796 | degree_sequence | False | | caffeine | 1.00e-08 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | caffeine | 1.00e-07 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | caffeine | 1.00e-06 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | @@ -18,71 +18,71 @@ | chlordimeform | 1.00e-07 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | chlordimeform | 1.00e-06 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | chlordimeform | 1.00e-05 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | -| chlorpyrifos_oxon | 1.00e-08 | 41.9 | topology-dependent | topology-dependent | 0.049 | degree_sequence | True | -| chlorpyrifos_oxon | 1.00e-07 | -4.453 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| chlorpyrifos_oxon | 1.00e-06 | -5.555 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| chlorpyrifos_oxon | 1.00e-05 | -5.550 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| clothianidin | 1.00e-08 | 0.100 | composition-dominated | composition-dominated | 0.545 | degree_sequence | False | -| clothianidin | 1.00e-07 | -4.774 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| clothianidin | 1.00e-06 | -6.173 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| clothianidin | 1.00e-05 | -6.173 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| ddt | 1.00e-08 | 0.003 | no-effect | no-effect | 0.904 | none_required | False | -| ddt | 1.00e-07 | 0.031 | no-effect | no-effect | 0.904 | none_required | False | -| ddt | 1.00e-06 | 0.308 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| ddt | 1.00e-05 | 2.982 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| deltamethrin | 1.00e-08 | 4.405 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| deltamethrin | 1.00e-07 | 24.0 | composition-dominated | composition-dominated | 0.507 | degree_sequence | False | -| deltamethrin | 1.00e-06 | 37.2 | composition-dominated | composition-dominated | 0.327 | degree_sequence | False | -| deltamethrin | 1.00e-05 | 38.2 | composition-dominated | composition-dominated | 0.396 | degree_sequence | False | +| chlorpyrifos_oxon | 1.00e-08 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| chlorpyrifos_oxon | 1.00e-07 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| chlorpyrifos_oxon | 1.00e-06 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| chlorpyrifos_oxon | 1.00e-05 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| clothianidin | 1.00e-08 | 0.100 | composition-dominated | composition-dominated | 0.656 | degree_sequence | False | +| clothianidin | 1.00e-07 | -4.774 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| clothianidin | 1.00e-06 | -6.173 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| clothianidin | 1.00e-05 | -6.173 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| ddt | 1.00e-08 | 0.003 | no-effect | no-effect | 1.000 | none_required | False | +| ddt | 1.00e-07 | 0.031 | no-effect | no-effect | 1.000 | none_required | False | +| ddt | 1.00e-06 | 0.308 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| ddt | 1.00e-05 | 2.982 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| deltamethrin | 1.00e-08 | 4.405 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| deltamethrin | 1.00e-07 | 24.0 | composition-dominated | composition-dominated | 0.557 | degree_sequence | False | +| deltamethrin | 1.00e-06 | 37.2 | composition-dominated | composition-dominated | 0.344 | degree_sequence | False | +| deltamethrin | 1.00e-05 | 38.2 | composition-dominated | composition-dominated | 0.426 | degree_sequence | False | | diazepam | 1.00e-08 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | diazepam | 1.00e-07 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | diazepam | 1.00e-06 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | diazepam | 1.00e-05 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | -| dieldrin | 1.00e-08 | 0.071 | topology-dependent | topology-dependent | 0.049 | degree_sequence | True | -| dieldrin | 1.00e-07 | 0.424 | topology-dependent | topology-dependent | 0.049 | degree_sequence | True | -| dieldrin | 1.00e-06 | 0.851 | topology-dependent | topology-dependent | 0.049 | degree_sequence | True | -| dieldrin | 1.00e-05 | 0.900 | composition-dominated | topology-dependent | 0.050 | degree_sequence | False | -| fipronil | 1.00e-08 | 0.168 | topology-dependent | topology-dependent | 0.049 | degree_sequence | True | -| fipronil | 1.00e-07 | 0.738 | topology-dependent | topology-dependent | 0.049 | degree_sequence | True | -| fipronil | 1.00e-06 | 0.900 | composition-dominated | topology-dependent | 0.050 | degree_sequence | False | -| fipronil | 1.00e-05 | 0.900 | composition-dominated | topology-dependent | 0.050 | degree_sequence | False | -| gaba | 1.00e-08 | -8.67e-07 | no-effect | no-effect | 0.049 | none_required | False | -| gaba | 1.00e-07 | -2.74e-05 | no-effect | no-effect | 0.049 | none_required | False | -| gaba | 1.00e-06 | -8.64e-04 | no-effect | no-effect | 0.049 | none_required | False | -| gaba | 1.00e-05 | -0.025 | no-effect | no-effect | 0.049 | none_required | False | -| imidacloprid | 1.00e-08 | -0.194 | composition-dominated | composition-dominated | 0.545 | degree_sequence | False | -| imidacloprid | 1.00e-07 | -5.680 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| imidacloprid | 1.00e-06 | -6.173 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| imidacloprid | 1.00e-05 | -6.173 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| ivermectin | 1.00e-08 | -0.016 | no-effect | no-effect | 0.065 | none_required | False | -| ivermectin | 1.00e-07 | -0.059 | no-effect | no-effect | 0.049 | none_required | False | -| ivermectin | 1.00e-06 | -0.184 | topology-dependent | topology-dependent | 0.049 | degree_sequence | True | -| ivermectin | 1.00e-05 | -0.297 | topology-dependent | topology-dependent | 0.049 | degree_sequence | True | -| nicotine | 1.00e-08 | 0.093 | composition-dominated | composition-dominated | 0.545 | degree_sequence | False | -| nicotine | 1.00e-07 | -0.260 | composition-dominated | composition-dominated | 0.545 | degree_sequence | False | -| nicotine | 1.00e-06 | -5.404 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| nicotine | 1.00e-05 | -6.173 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| nitenpyram | 1.00e-08 | 0.191 | composition-dominated | composition-dominated | 0.545 | degree_sequence | False | -| nitenpyram | 1.00e-07 | -3.378 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| nitenpyram | 1.00e-06 | -6.173 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| nitenpyram | 1.00e-05 | -6.173 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| permethrin | 1.00e-08 | 0.500 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| permethrin | 1.00e-07 | 4.710 | composition-dominated | composition-dominated | 0.904 | degree_sequence | False | -| permethrin | 1.00e-06 | 24.8 | composition-dominated | composition-dominated | 0.462 | degree_sequence | False | -| permethrin | 1.00e-05 | 37.2 | composition-dominated | composition-dominated | 0.327 | degree_sequence | False | -| picrotoxin | 1.00e-08 | 0.008 | no-effect | no-effect | 0.049 | none_required | False | -| picrotoxin | 1.00e-07 | 0.071 | topology-dependent | topology-dependent | 0.049 | degree_sequence | True | -| picrotoxin | 1.00e-06 | 0.424 | topology-dependent | topology-dependent | 0.049 | degree_sequence | True | -| picrotoxin | 1.00e-05 | 0.851 | topology-dependent | topology-dependent | 0.049 | degree_sequence | True | -| spinosad | 1.00e-08 | 0.006 | no-effect | no-effect | 0.545 | none_required | False | -| spinosad | 1.00e-07 | 0.055 | no-effect | no-effect | 0.545 | none_required | False | -| spinosad | 1.00e-06 | 0.170 | composition-dominated | composition-dominated | 0.545 | degree_sequence | False | -| spinosad | 1.00e-05 | -3.041 | composition-dominated | composition-dominated | 0.891 | degree_sequence | False | -| sulfoxaflor | 1.00e-08 | 2.02e-04 | no-effect | no-effect | 0.545 | none_required | False | -| sulfoxaflor | 1.00e-07 | 0.002 | no-effect | no-effect | 0.545 | none_required | False | -| sulfoxaflor | 1.00e-06 | 0.020 | no-effect | no-effect | 0.545 | none_required | False | -| sulfoxaflor | 1.00e-05 | 0.143 | composition-dominated | composition-dominated | 0.545 | degree_sequence | False | -| thiamethoxam | 1.00e-08 | 3.03e-04 | no-effect | no-effect | 0.545 | none_required | False | -| thiamethoxam | 1.00e-07 | 0.003 | no-effect | no-effect | 0.545 | none_required | False | -| thiamethoxam | 1.00e-06 | 0.029 | no-effect | no-effect | 0.545 | none_required | False | -| thiamethoxam | 1.00e-05 | 0.177 | composition-dominated | composition-dominated | 0.545 | degree_sequence | False | +| dieldrin | 1.00e-08 | 0.071 | composition-dominated | topology-dependent | 0.062 | degree_sequence | False | +| dieldrin | 1.00e-07 | 0.424 | composition-dominated | topology-dependent | 0.062 | degree_sequence | False | +| dieldrin | 1.00e-06 | 0.851 | composition-dominated | topology-dependent | 0.062 | degree_sequence | False | +| dieldrin | 1.00e-05 | 0.900 | composition-dominated | topology-dependent | 0.062 | degree_sequence | False | +| fipronil | 1.00e-08 | 0.168 | composition-dominated | topology-dependent | 0.062 | degree_sequence | False | +| fipronil | 1.00e-07 | 0.738 | composition-dominated | topology-dependent | 0.062 | degree_sequence | False | +| fipronil | 1.00e-06 | 0.900 | composition-dominated | topology-dependent | 0.062 | degree_sequence | False | +| fipronil | 1.00e-05 | 0.900 | composition-dominated | topology-dependent | 0.062 | degree_sequence | False | +| gaba | 1.00e-08 | -8.67e-07 | no-effect | no-effect | 0.062 | none_required | False | +| gaba | 1.00e-07 | -2.74e-05 | no-effect | no-effect | 0.062 | none_required | False | +| gaba | 1.00e-06 | -8.64e-04 | no-effect | no-effect | 0.062 | none_required | False | +| gaba | 1.00e-05 | -0.025 | no-effect | no-effect | 0.062 | none_required | False | +| imidacloprid | 1.00e-08 | -0.194 | composition-dominated | composition-dominated | 0.656 | degree_sequence | False | +| imidacloprid | 1.00e-07 | -5.680 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| imidacloprid | 1.00e-06 | -6.173 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| imidacloprid | 1.00e-05 | -6.173 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| ivermectin | 1.00e-08 | -0.013 | no-effect | no-effect | 0.071 | none_required | False | +| ivermectin | 1.00e-07 | -0.031 | no-effect | no-effect | 0.071 | none_required | False | +| ivermectin | 1.00e-06 | -0.037 | no-effect | no-effect | 0.071 | none_required | False | +| ivermectin | 1.00e-05 | -0.037 | no-effect | no-effect | 0.071 | none_required | False | +| nicotine | 1.00e-08 | 0.093 | composition-dominated | composition-dominated | 0.656 | degree_sequence | False | +| nicotine | 1.00e-07 | -0.260 | composition-dominated | composition-dominated | 0.656 | degree_sequence | False | +| nicotine | 1.00e-06 | -5.404 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| nicotine | 1.00e-05 | -6.173 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| nitenpyram | 1.00e-08 | 0.191 | composition-dominated | composition-dominated | 0.656 | degree_sequence | False | +| nitenpyram | 1.00e-07 | -3.378 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| nitenpyram | 1.00e-06 | -6.173 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| nitenpyram | 1.00e-05 | -6.173 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| permethrin | 1.00e-08 | 0.500 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| permethrin | 1.00e-07 | 4.710 | composition-dominated | composition-dominated | 1.000 | degree_sequence | False | +| permethrin | 1.00e-06 | 24.8 | composition-dominated | composition-dominated | 0.508 | degree_sequence | False | +| permethrin | 1.00e-05 | 37.2 | composition-dominated | composition-dominated | 0.344 | degree_sequence | False | +| picrotoxin | 1.00e-08 | 0.008 | no-effect | no-effect | 0.062 | none_required | False | +| picrotoxin | 1.00e-07 | 0.071 | composition-dominated | topology-dependent | 0.062 | degree_sequence | False | +| picrotoxin | 1.00e-06 | 0.424 | composition-dominated | topology-dependent | 0.062 | degree_sequence | False | +| picrotoxin | 1.00e-05 | 0.851 | composition-dominated | topology-dependent | 0.062 | degree_sequence | False | +| spinosad | 1.00e-08 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| spinosad | 1.00e-07 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| spinosad | 1.00e-06 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| spinosad | 1.00e-05 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| sulfoxaflor | 1.00e-08 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| sulfoxaflor | 1.00e-07 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| sulfoxaflor | 1.00e-06 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| sulfoxaflor | 1.00e-05 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| thiamethoxam | 1.00e-08 | 3.03e-04 | no-effect | no-effect | 0.656 | none_required | False | +| thiamethoxam | 1.00e-07 | 0.003 | no-effect | no-effect | 0.656 | none_required | False | +| thiamethoxam | 1.00e-06 | 0.029 | no-effect | no-effect | 0.656 | none_required | False | +| thiamethoxam | 1.00e-05 | 0.177 | composition-dominated | composition-dominated | 0.656 | degree_sequence | False | diff --git a/papers/tables/T12b_dependence_landscape_taste_motor.csv b/papers/tables/T12b_dependence_landscape_taste_motor.csv index 4c221f3..920a4e0 100644 --- a/papers/tables/T12b_dependence_landscape_taste_motor.csv +++ b/papers/tables/T12b_dependence_landscape_taste_motor.csv @@ -1,12 +1,12 @@ compound,conc_M,real_effect,vehicle,effect_floor,delta,class,class_raw,class_absolute_floor,q_value,necessary_level,necessary_level_verdict,non_monotone,p_sign,p_sign_wm,p_weight,p_degree,p_ER,q_weight,q_degree,verdict_weight,verdict_degree -acetamiprid,1e-08,0.45339694651465123,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9302325581395349,0.6744186046511628,0.28903654485049834,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -acetamiprid,1e-07,-7.375166864803854,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.2292358803986711,0.9501661129568106,0.14950166112956811,0.0033222591362126247,0.0033222591362126247,0.22832980972515857,0.006489994591671174,indeterminate,distinguishable -acetamiprid,1e-06,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.11491108071135431,0.006489994591671174,indeterminate,distinguishable -acetamiprid,1e-05,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.11491108071135431,0.006489994591671174,indeterminate,distinguishable -acetylcholine,1e-08,0.001799501517826485,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.893687707641196,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -acetylcholine,1e-07,0.028010151444085807,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.8970099667774086,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -acetylcholine,1e-06,0.32733373835234936,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9169435215946844,0.6810631229235881,0.2857142857142857,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -acetylcholine,1e-05,-3.2993138264269763,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.6312292358803987,0.813953488372093,0.2159468438538206,0.0033222591362126247,0.0033222591362126247,0.3239202657807309,0.006489994591671174,indeterminate,distinguishable +acetamiprid,1e-08,0.45339694651465123,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9302325581395349,0.6744186046511628,0.28903654485049834,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable +acetamiprid,1e-07,-7.375166864803854,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.2292358803986711,0.9501661129568106,0.14950166112956811,0.0033222591362126247,0.0033222591362126247,0.2730030333670374,0.008088978766430738,indeterminate,distinguishable +acetamiprid,1e-06,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.1347233360064154,0.008088978766430738,indeterminate,distinguishable +acetamiprid,1e-05,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.1347233360064154,0.008088978766430738,indeterminate,distinguishable +acetylcholine,1e-08,0.001799501517826485,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.008088978766430738,none_required,,False,0.893687707641196,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable +acetylcholine,1e-07,0.028010151444085807,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.008088978766430738,none_required,,False,0.8970099667774086,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable +acetylcholine,1e-06,0.32733373835234936,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9169435215946844,0.6810631229235881,0.2857142857142857,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable +acetylcholine,1e-05,-3.2993138264269763,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.6312292358803987,0.813953488372093,0.2159468438538206,0.0033222591362126247,0.0033222591362126247,0.3859475507174666,0.008088978766430738,indeterminate,distinguishable caffeine,1e-08,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance caffeine,1e-07,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance caffeine,1e-06,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance @@ -15,71 +15,71 @@ chlordimeform,1e-08,0.0,14.345707408485804,0.14345707408485806,0.717285370424290 chlordimeform,1e-07,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance chlordimeform,1e-06,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance chlordimeform,1e-05,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance -chlorpyrifos_oxon,1e-08,113.02014979865878,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.006644518272425249,0.09302325581395349,0.34219269102990035,0.0033222591362126247,0.0033222591362126247,0.4106312292358804,0.006489994591671174,indeterminate,distinguishable -chlorpyrifos_oxon,1e-07,-9.400518990661134,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.1362126245847176,0.9900332225913622,0.11295681063122924,0.0033222591362126247,0.0033222591362126247,0.1757105943152455,0.006489994591671174,indeterminate,distinguishable -chlorpyrifos_oxon,1e-06,-11.274548756961796,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.07641196013289037,0.9235880398671097,0.07973421926910298,0.0033222591362126247,0.0033222591362126247,0.12637121544537078,0.006489994591671174,indeterminate,distinguishable -chlorpyrifos_oxon,1e-05,-11.265710474999171,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.07641196013289037,0.9235880398671097,0.07973421926910298,0.0033222591362126247,0.0033222591362126247,0.12637121544537078,0.006489994591671174,indeterminate,distinguishable -clothianidin,1e-08,0.23703647392069982,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9069767441860465,0.6810631229235881,0.2857142857142857,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -clothianidin,1e-07,-9.966794607205912,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.11627906976744186,0.9601328903654485,0.09966777408637874,0.0033222591362126247,0.0033222591362126247,0.15648772006085634,0.006489994591671174,indeterminate,distinguishable -clothianidin,1e-06,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.11491108071135431,0.006489994591671174,indeterminate,distinguishable -clothianidin,1e-05,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.11491108071135431,0.006489994591671174,indeterminate,distinguishable -ddt,1e-08,0.006558808143186212,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.6112956810631229,0.7574750830564784,0.30564784053156147,0.0033222591362126247,0.0033222591362126247,0.3761375126390294,0.006489994591671174,equivalent_within_tolerance,distinguishable -ddt,1e-07,0.06556257274796096,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.6212624584717608,0.7574750830564784,0.3089700996677741,0.0033222591362126247,0.0033222591362126247,0.3761375126390294,0.006489994591671174,equivalent_within_tolerance,distinguishable -ddt,1e-06,0.6528758087588304,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,equivalent_within_tolerance,False,0.6578073089700996,0.7408637873754153,0.3122923588039867,0.0033222591362126247,0.0033222591362126247,0.3774468797055379,0.006489994591671174,equivalent_within_tolerance,distinguishable -ddt,1e-05,6.237556740372629,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.9767441860465116,0.6146179401993356,0.3687707641196013,0.0033222591362126247,0.0033222591362126247,0.4362921716344579,0.006489994591671174,indeterminate,distinguishable -deltamethrin,1e-08,9.288999514164995,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.9501661129568106,0.5880398671096345,0.4186046511627907,0.0033222591362126247,0.0033222591362126247,0.4917872824849569,0.006489994591671174,indeterminate,distinguishable -deltamethrin,1e-07,29.104258377410016,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,indeterminate,False,0.12624584717607973,0.08305647840531562,0.0033222591362126247,0.023255813953488372,0.0033222591362126247,0.006489994591671174,0.042467138523761376,distinguishable,distinguishable -deltamethrin,1e-06,54.14423390901861,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.012542461458061144,topology_without_weight_pairing,indeterminate,False,0.11627906976744186,0.18272425249169436,0.2823920265780731,0.006644518272425249,0.0033222591362126247,0.35968992248062015,0.012542461458061144,indeterminate,distinguishable -deltamethrin,1e-05,54.80105342774155,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.012542461458061144,topology_without_weight_pairing,indeterminate,False,0.09302325581395349,0.15282392026578073,0.2724252491694352,0.006644518272425249,0.0033222591362126247,0.35968992248062015,0.012542461458061144,indeterminate,distinguishable +chlorpyrifos_oxon,1e-08,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +chlorpyrifos_oxon,1e-07,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +chlorpyrifos_oxon,1e-06,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +chlorpyrifos_oxon,1e-05,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +clothianidin,1e-08,0.23703647392069982,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9069767441860465,0.6810631229235881,0.2857142857142857,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable +clothianidin,1e-07,-9.966794607205912,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.11627906976744186,0.9601328903654485,0.09966777408637874,0.0033222591362126247,0.0033222591362126247,0.186046511627907,0.008088978766430738,indeterminate,distinguishable +clothianidin,1e-06,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.1347233360064154,0.008088978766430738,indeterminate,distinguishable +clothianidin,1e-05,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.1347233360064154,0.008088978766430738,indeterminate,distinguishable +ddt,1e-08,0.006558808143186212,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.008088978766430738,none_required,,False,0.6112956810631229,0.7574750830564784,0.30564784053156147,0.0033222591362126247,0.0033222591362126247,0.4474739374498798,0.008088978766430738,equivalent_within_tolerance,distinguishable +ddt,1e-07,0.06556257274796096,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.008088978766430738,none_required,,False,0.6212624584717608,0.7574750830564784,0.3089700996677741,0.0033222591362126247,0.0033222591362126247,0.4474739374498798,0.008088978766430738,equivalent_within_tolerance,distinguishable +ddt,1e-06,0.6528758087588304,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,equivalent_within_tolerance,False,0.6578073089700996,0.7408637873754153,0.3122923588039867,0.0033222591362126247,0.0033222591362126247,0.4484197972570066,0.008088978766430738,equivalent_within_tolerance,distinguishable +ddt,1e-05,6.237556740372629,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.9767441860465116,0.6146179401993356,0.3687707641196013,0.0033222591362126247,0.0033222591362126247,0.5250295624753646,0.008088978766430738,indeterminate,distinguishable +deltamethrin,1e-08,9.288999514164995,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.9501661129568106,0.5880398671096345,0.4186046511627907,0.0033222591362126247,0.0033222591362126247,0.5909712722298222,0.008088978766430738,indeterminate,distinguishable +deltamethrin,1e-07,29.104258377410016,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,degree_sequence,indeterminate,True,0.12624584717607973,0.08305647840531562,0.0033222591362126247,0.023255813953488372,0.0033222591362126247,0.008088978766430738,0.05008944543828264,distinguishable,indeterminate +deltamethrin,1e-06,54.14423390901861,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.015503875968992248,topology_without_weight_pairing,indeterminate,False,0.11627906976744186,0.18272425249169436,0.2823920265780731,0.006644518272425249,0.0033222591362126247,0.42971804898127186,0.015503875968992248,indeterminate,distinguishable +deltamethrin,1e-05,54.80105342774155,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.015503875968992248,topology_without_weight_pairing,indeterminate,False,0.09302325581395349,0.15282392026578073,0.2724252491694352,0.006644518272425249,0.0033222591362126247,0.42971804898127186,0.015503875968992248,indeterminate,distinguishable diazepam,1e-08,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance diazepam,1e-07,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance diazepam,1e-06,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance diazepam,1e-05,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance -dieldrin,1e-08,0.356292538854051,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.292358803986711,0.6046511627906976,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -dieldrin,1e-07,2.358455281893889,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.37209302325581395,0.7275747508305648,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -dieldrin,1e-06,5.516203307325238,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.4850498338870432,0.7973421926910299,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -dieldrin,1e-05,5.944493581459351,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.4950166112956811,0.8006644518272426,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -fipronil,1e-08,0.8698058843710257,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.3289036544850498,0.6312292358803987,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -fipronil,1e-07,4.574334628338235,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.46511627906976744,0.7873754152823921,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -fipronil,1e-06,5.944493581459351,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.4950166112956811,0.8006644518272426,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -fipronil,1e-05,5.944493581459351,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.4950166112956811,0.8006644518272426,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -gaba,1e-08,-4.269749503293951e-06,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.27906976744186046,0.5980066445182725,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -gaba,1e-07,-0.00013500781989073118,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.27906976744186046,0.5980066445182725,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -gaba,1e-06,-0.004255851246456999,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.2823920265780731,0.5980066445182725,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -gaba,1e-05,-0.12229139007903278,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.2857142857142857,0.5913621262458472,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -imidacloprid,1e-08,-0.45812824345463277,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,equivalent_within_tolerance,False,0.8737541528239202,0.7009966777408638,0.26578073089701,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -imidacloprid,1e-07,-11.47357179820895,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.07641196013289037,0.9136212624584718,0.07641196013289037,0.0033222591362126247,0.0033222591362126247,0.12463310002257846,0.006489994591671174,indeterminate,distinguishable -imidacloprid,1e-06,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.11491108071135431,0.006489994591671174,indeterminate,distinguishable -imidacloprid,1e-05,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.11491108071135431,0.006489994591671174,indeterminate,distinguishable -ivermectin,1e-08,-0.03484237201594986,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,composition-dominated,0.06601650412603151,none_required,,False,0.292358803986711,0.3488372093023256,0.036544850498338874,0.3521594684385382,0.6644518272425249,0.06601650412603151,0.41959426026719443,equivalent_within_tolerance,equivalent_within_tolerance -ivermectin,1e-07,-0.17961943494856492,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.5083056478405316,0.4418604651162791,0.009966777408637873,0.0033222591362126247,0.0033222591362126247,0.018604651162790697,0.006489994591671174,distinguishable,distinguishable -ivermectin,1e-06,-0.7435498908369098,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.7807308970099668,0.8338870431893688,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -ivermectin,1e-05,-1.2358102786765368,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.5249169435215947,0.9102990033222591,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -nicotine,1e-08,0.22089178394916686,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9069767441860465,0.6810631229235881,0.2857142857142857,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -nicotine,1e-07,-0.6120735745119354,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,equivalent_within_tolerance,False,0.867109634551495,0.7043189368770764,0.2558139534883721,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -nicotine,1e-06,-11.029611556363646,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.07641196013289037,0.9302325581395349,0.07973421926910298,0.0033222591362126247,0.0033222591362126247,0.12637121544537078,0.006489994591671174,indeterminate,distinguishable -nicotine,1e-05,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.11491108071135431,0.006489994591671174,indeterminate,distinguishable -nitenpyram,1e-08,0.45339694651465123,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9302325581395349,0.6744186046511628,0.28903654485049834,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -nitenpyram,1e-07,-7.375166864803854,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.2292358803986711,0.9501661129568106,0.14950166112956811,0.0033222591362126247,0.0033222591362126247,0.22832980972515857,0.006489994591671174,indeterminate,distinguishable -nitenpyram,1e-06,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.11491108071135431,0.006489994591671174,indeterminate,distinguishable -nitenpyram,1e-05,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.11491108071135431,0.006489994591671174,indeterminate,distinguishable -permethrin,1e-08,1.0577092189958517,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,equivalent_within_tolerance,False,0.6910299003322259,0.7242524916943521,0.3089700996677741,0.0033222591362126247,0.0033222591362126247,0.3761375126390294,0.006489994591671174,equivalent_within_tolerance,distinguishable -permethrin,1e-07,9.983088432977482,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.946843853820598,0.584717607973422,0.44518272425249167,0.0033222591362126247,0.0033222591362126247,0.5193798449612402,0.006489994591671174,indeterminate,distinguishable -permethrin,1e-06,29.409204580561596,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,indeterminate,False,0.11627906976744186,0.08305647840531562,0.0033222591362126247,0.019933554817275746,0.0033222591362126247,0.006489994591671174,0.03680040889343215,distinguishable,distinguishable -permethrin,1e-05,54.19622715044788,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.012542461458061144,topology_without_weight_pairing,indeterminate,False,0.11295681063122924,0.18272425249169436,0.2823920265780731,0.006644518272425249,0.0033222591362126247,0.35968992248062015,0.012542461458061144,indeterminate,distinguishable -picrotoxin,1e-08,0.03751262429897828,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.2823920265780731,0.5980066445182725,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -picrotoxin,1e-07,0.356292538854051,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.292358803986711,0.6046511627906976,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -picrotoxin,1e-06,2.358455281893889,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.37209302325581395,0.7275747508305648,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -picrotoxin,1e-05,5.516203307325238,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.4850498338870432,0.7973421926910299,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.006489994591671174,0.006489994591671174,distinguishable,distinguishable -spinosad,1e-08,0.014207395519038712,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.8970099667774086,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -spinosad,1e-07,0.13003485133672932,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.9003322259136213,0.6843853820598007,0.2857142857142857,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -spinosad,1e-06,0.4025585847551074,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9235880398671097,0.6810631229235881,0.2857142857142857,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -spinosad,1e-05,-6.703823427435451,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,indeterminate,False,0.2724252491694352,0.9302325581395349,0.15946843853820597,0.0033222591362126247,0.0033222591362126247,0.2413576367064739,0.006489994591671174,indeterminate,distinguishable -sulfoxaflor,1e-08,0.00047802132707097655,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.893687707641196,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -sulfoxaflor,1e-07,0.0047664016996886716,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.893687707641196,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -sulfoxaflor,1e-06,0.046296506580789654,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.8970099667774086,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -sulfoxaflor,1e-05,0.33910898339945206,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9169435215946844,0.6810631229235881,0.2857142857142857,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -thiamethoxam,1e-08,0.0007169167960228151,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.893687707641196,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -thiamethoxam,1e-07,0.007138109061877174,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.8970099667774086,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -thiamethoxam,1e-06,0.06832111891335835,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.006489994591671174,none_required,,False,0.8970099667774086,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable -thiamethoxam,1e-05,0.41937492441378055,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.006489994591671174,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9269102990033222,0.6810631229235881,0.2857142857142857,0.0033222591362126247,0.0033222591362126247,0.35968992248062015,0.006489994591671174,equivalent_within_tolerance,distinguishable +dieldrin,1e-08,0.356292538854051,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.292358803986711,0.6046511627906976,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +dieldrin,1e-07,2.358455281893889,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.37209302325581395,0.7275747508305648,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +dieldrin,1e-06,5.516203307325238,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.4850498338870432,0.7973421926910299,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +dieldrin,1e-05,5.944493581459351,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.4950166112956811,0.8006644518272426,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +fipronil,1e-08,0.8698058843710257,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.3289036544850498,0.6312292358803987,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +fipronil,1e-07,4.574334628338235,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.46511627906976744,0.7873754152823921,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +fipronil,1e-06,5.944493581459351,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.4950166112956811,0.8006644518272426,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +fipronil,1e-05,5.944493581459351,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.4950166112956811,0.8006644518272426,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +gaba,1e-08,-4.269749503293951e-06,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.008088978766430738,none_required,,False,0.27906976744186046,0.5980066445182725,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +gaba,1e-07,-0.00013500781989073118,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.008088978766430738,none_required,,False,0.27906976744186046,0.5980066445182725,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +gaba,1e-06,-0.004255851246456999,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.008088978766430738,none_required,,False,0.2823920265780731,0.5980066445182725,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +gaba,1e-05,-0.12229139007903278,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.008088978766430738,none_required,,False,0.2857142857142857,0.5913621262458472,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +imidacloprid,1e-08,-0.45812824345463277,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,equivalent_within_tolerance,False,0.8737541528239202,0.7009966777408638,0.26578073089701,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable +imidacloprid,1e-07,-11.47357179820895,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.07641196013289037,0.9136212624584718,0.07641196013289037,0.0033222591362126247,0.0033222591362126247,0.14587737843551796,0.008088978766430738,indeterminate,distinguishable +imidacloprid,1e-06,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.1347233360064154,0.008088978766430738,indeterminate,distinguishable +imidacloprid,1e-05,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.1347233360064154,0.008088978766430738,indeterminate,distinguishable +ivermectin,1e-08,-0.019972099842439306,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.04349139232860163,none_required,,False,0.24584717607973422,0.30564784053156147,0.19269102990033224,0.019933554817275746,0.03322259136212625,0.3480870217554389,0.04349139232860163,equivalent_within_tolerance,distinguishable +ivermectin,1e-07,-0.04540873117066013,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.022325581395348834,none_required,,False,0.21926910299003322,0.29900332225913623,0.24584717607973422,0.009966777408637873,0.023255813953488372,0.42971804898127186,0.022325581395348834,equivalent_within_tolerance,distinguishable +ivermectin,1e-06,-0.05234164208322589,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.022325581395348834,none_required,,False,0.21926910299003322,0.3023255813953488,0.2591362126245847,0.009966777408637873,0.023255813953488372,0.42971804898127186,0.022325581395348834,equivalent_within_tolerance,distinguishable +ivermectin,1e-05,-0.05316580295906803,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.022325581395348834,none_required,,False,0.21926910299003322,0.3023255813953488,0.2591362126245847,0.009966777408637873,0.019933554817275746,0.42971804898127186,0.022325581395348834,equivalent_within_tolerance,distinguishable +nicotine,1e-08,0.22089178394916686,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9069767441860465,0.6810631229235881,0.2857142857142857,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable +nicotine,1e-07,-0.6120735745119354,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,equivalent_within_tolerance,False,0.867109634551495,0.7043189368770764,0.2558139534883721,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable +nicotine,1e-06,-11.029611556363646,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.07641196013289037,0.9302325581395349,0.07973421926910298,0.0033222591362126247,0.0033222591362126247,0.1505095374967337,0.008088978766430738,indeterminate,distinguishable +nicotine,1e-05,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.1347233360064154,0.008088978766430738,indeterminate,distinguishable +nitenpyram,1e-08,0.45339694651465123,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9302325581395349,0.6744186046511628,0.28903654485049834,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable +nitenpyram,1e-07,-7.375166864803854,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.2292358803986711,0.9501661129568106,0.14950166112956811,0.0033222591362126247,0.0033222591362126247,0.2730030333670374,0.008088978766430738,indeterminate,distinguishable +nitenpyram,1e-06,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.1347233360064154,0.008088978766430738,indeterminate,distinguishable +nitenpyram,1e-05,-12.23073546730731,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.06312292358803986,0.8903654485049833,0.06976744186046512,0.0033222591362126247,0.0033222591362126247,0.1347233360064154,0.008088978766430738,indeterminate,distinguishable +permethrin,1e-08,1.0577092189958517,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,equivalent_within_tolerance,False,0.6910299003322259,0.7242524916943521,0.3089700996677741,0.0033222591362126247,0.0033222591362126247,0.4474739374498798,0.008088978766430738,equivalent_within_tolerance,distinguishable +permethrin,1e-07,9.983088432977482,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,indeterminate,False,0.946843853820598,0.584717607973422,0.44518272425249167,0.0033222591362126247,0.0033222591362126247,0.6232558139534883,0.008088978766430738,indeterminate,distinguishable +permethrin,1e-06,29.409204580561596,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,wiring_with_weighted_transmitter_balance,indeterminate,False,0.11627906976744186,0.08305647840531562,0.0033222591362126247,0.019933554817275746,0.0033222591362126247,0.008088978766430738,0.04349139232860163,distinguishable,distinguishable +permethrin,1e-05,54.19622715044788,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.015503875968992248,topology_without_weight_pairing,indeterminate,False,0.11295681063122924,0.18272425249169436,0.2823920265780731,0.006644518272425249,0.0033222591362126247,0.42971804898127186,0.015503875968992248,indeterminate,distinguishable +picrotoxin,1e-08,0.03751262429897828,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.008088978766430738,none_required,,False,0.2823920265780731,0.5980066445182725,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +picrotoxin,1e-07,0.356292538854051,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.292358803986711,0.6046511627906976,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +picrotoxin,1e-06,2.358455281893889,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.37209302325581395,0.7275747508305648,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +picrotoxin,1e-05,5.516203307325238,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,wiring_with_weighted_transmitter_balance,equivalent_within_tolerance,False,0.4850498338870432,0.7973421926910299,0.0033222591362126247,0.0033222591362126247,0.0033222591362126247,0.008088978766430738,0.008088978766430738,distinguishable,distinguishable +spinosad,1e-08,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +spinosad,1e-07,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +spinosad,1e-06,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +spinosad,1e-05,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +sulfoxaflor,1e-08,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +sulfoxaflor,1e-07,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +sulfoxaflor,1e-06,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +sulfoxaflor,1e-05,0.0,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,no-effect,1.0,none_required,,False,1.0,1.0,1.0,1.0,1.0,1.0,1.0,equivalent_within_tolerance,equivalent_within_tolerance +thiamethoxam,1e-08,0.0007169167960228151,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.008088978766430738,none_required,,False,0.893687707641196,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable +thiamethoxam,1e-07,0.007138109061877174,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.008088978766430738,none_required,,False,0.8970099667774086,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable +thiamethoxam,1e-06,0.06832111891335835,14.345707408485804,0.14345707408485806,0.7172853704242903,no-effect,no-effect,topology-dependent,0.008088978766430738,none_required,,False,0.8970099667774086,0.6843853820598007,0.2823920265780731,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable +thiamethoxam,1e-05,0.41937492441378055,14.345707408485804,0.14345707408485806,0.7172853704242903,topology-dependent,topology-dependent,topology-dependent,0.008088978766430738,topology_without_weight_pairing,equivalent_within_tolerance,False,0.9269102990033222,0.6810631229235881,0.2857142857142857,0.0033222591362126247,0.0033222591362126247,0.42971804898127186,0.008088978766430738,equivalent_within_tolerance,distinguishable diff --git a/papers/tables/T12b_dependence_landscape_taste_motor.md b/papers/tables/T12b_dependence_landscape_taste_motor.md index 8a9b24e..92406f6 100644 --- a/papers/tables/T12b_dependence_landscape_taste_motor.md +++ b/papers/tables/T12b_dependence_landscape_taste_motor.md @@ -2,14 +2,14 @@ | compound | conc_M | real_effect | class | class_raw | q_value | necessary_level | non_monotone | |---|---|---|---|---|---|---|---| -| acetamiprid | 1.00e-08 | 0.453 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| acetamiprid | 1.00e-07 | -7.375 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| acetamiprid | 1.00e-06 | -12.2 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| acetamiprid | 1.00e-05 | -12.2 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| acetylcholine | 1.00e-08 | 0.002 | no-effect | no-effect | 0.006 | none_required | False | -| acetylcholine | 1.00e-07 | 0.028 | no-effect | no-effect | 0.006 | none_required | False | -| acetylcholine | 1.00e-06 | 0.327 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| acetylcholine | 1.00e-05 | -3.299 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | +| acetamiprid | 1.00e-08 | 0.453 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| acetamiprid | 1.00e-07 | -7.375 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| acetamiprid | 1.00e-06 | -12.2 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| acetamiprid | 1.00e-05 | -12.2 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| acetylcholine | 1.00e-08 | 0.002 | no-effect | no-effect | 0.008 | none_required | False | +| acetylcholine | 1.00e-07 | 0.028 | no-effect | no-effect | 0.008 | none_required | False | +| acetylcholine | 1.00e-06 | 0.327 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| acetylcholine | 1.00e-05 | -3.299 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | | caffeine | 1.00e-08 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | caffeine | 1.00e-07 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | caffeine | 1.00e-06 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | @@ -18,71 +18,71 @@ | chlordimeform | 1.00e-07 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | chlordimeform | 1.00e-06 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | chlordimeform | 1.00e-05 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | -| chlorpyrifos_oxon | 1.00e-08 | 113 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| chlorpyrifos_oxon | 1.00e-07 | -9.401 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| chlorpyrifos_oxon | 1.00e-06 | -11.3 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| chlorpyrifos_oxon | 1.00e-05 | -11.3 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| clothianidin | 1.00e-08 | 0.237 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| clothianidin | 1.00e-07 | -9.967 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| clothianidin | 1.00e-06 | -12.2 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| clothianidin | 1.00e-05 | -12.2 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| ddt | 1.00e-08 | 0.007 | no-effect | no-effect | 0.006 | none_required | False | -| ddt | 1.00e-07 | 0.066 | no-effect | no-effect | 0.006 | none_required | False | -| ddt | 1.00e-06 | 0.653 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| ddt | 1.00e-05 | 6.238 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| deltamethrin | 1.00e-08 | 9.289 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| deltamethrin | 1.00e-07 | 29.1 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| deltamethrin | 1.00e-06 | 54.1 | topology-dependent | topology-dependent | 0.013 | topology_without_weight_pairing | False | -| deltamethrin | 1.00e-05 | 54.8 | topology-dependent | topology-dependent | 0.013 | topology_without_weight_pairing | False | +| chlorpyrifos_oxon | 1.00e-08 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| chlorpyrifos_oxon | 1.00e-07 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| chlorpyrifos_oxon | 1.00e-06 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| chlorpyrifos_oxon | 1.00e-05 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| clothianidin | 1.00e-08 | 0.237 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| clothianidin | 1.00e-07 | -9.967 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| clothianidin | 1.00e-06 | -12.2 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| clothianidin | 1.00e-05 | -12.2 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| ddt | 1.00e-08 | 0.007 | no-effect | no-effect | 0.008 | none_required | False | +| ddt | 1.00e-07 | 0.066 | no-effect | no-effect | 0.008 | none_required | False | +| ddt | 1.00e-06 | 0.653 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| ddt | 1.00e-05 | 6.238 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| deltamethrin | 1.00e-08 | 9.289 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| deltamethrin | 1.00e-07 | 29.1 | topology-dependent | topology-dependent | 0.008 | degree_sequence | True | +| deltamethrin | 1.00e-06 | 54.1 | topology-dependent | topology-dependent | 0.016 | topology_without_weight_pairing | False | +| deltamethrin | 1.00e-05 | 54.8 | topology-dependent | topology-dependent | 0.016 | topology_without_weight_pairing | False | | diazepam | 1.00e-08 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | diazepam | 1.00e-07 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | diazepam | 1.00e-06 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | | diazepam | 1.00e-05 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | -| dieldrin | 1.00e-08 | 0.356 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| dieldrin | 1.00e-07 | 2.358 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| dieldrin | 1.00e-06 | 5.516 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| dieldrin | 1.00e-05 | 5.944 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| fipronil | 1.00e-08 | 0.870 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| fipronil | 1.00e-07 | 4.574 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| fipronil | 1.00e-06 | 5.944 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| fipronil | 1.00e-05 | 5.944 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| gaba | 1.00e-08 | -4.27e-06 | no-effect | no-effect | 0.006 | none_required | False | -| gaba | 1.00e-07 | -1.35e-04 | no-effect | no-effect | 0.006 | none_required | False | -| gaba | 1.00e-06 | -0.004 | no-effect | no-effect | 0.006 | none_required | False | -| gaba | 1.00e-05 | -0.122 | no-effect | no-effect | 0.006 | none_required | False | -| imidacloprid | 1.00e-08 | -0.458 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| imidacloprid | 1.00e-07 | -11.5 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| imidacloprid | 1.00e-06 | -12.2 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| imidacloprid | 1.00e-05 | -12.2 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| ivermectin | 1.00e-08 | -0.035 | no-effect | no-effect | 0.066 | none_required | False | -| ivermectin | 1.00e-07 | -0.180 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| ivermectin | 1.00e-06 | -0.744 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| ivermectin | 1.00e-05 | -1.236 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| nicotine | 1.00e-08 | 0.221 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| nicotine | 1.00e-07 | -0.612 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| nicotine | 1.00e-06 | -11.0 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| nicotine | 1.00e-05 | -12.2 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| nitenpyram | 1.00e-08 | 0.453 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| nitenpyram | 1.00e-07 | -7.375 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| nitenpyram | 1.00e-06 | -12.2 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| nitenpyram | 1.00e-05 | -12.2 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| permethrin | 1.00e-08 | 1.058 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| permethrin | 1.00e-07 | 9.983 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| permethrin | 1.00e-06 | 29.4 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| permethrin | 1.00e-05 | 54.2 | topology-dependent | topology-dependent | 0.013 | topology_without_weight_pairing | False | -| picrotoxin | 1.00e-08 | 0.038 | no-effect | no-effect | 0.006 | none_required | False | -| picrotoxin | 1.00e-07 | 0.356 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| picrotoxin | 1.00e-06 | 2.358 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| picrotoxin | 1.00e-05 | 5.516 | topology-dependent | topology-dependent | 0.006 | wiring_with_weighted_transmitter_balance | False | -| spinosad | 1.00e-08 | 0.014 | no-effect | no-effect | 0.006 | none_required | False | -| spinosad | 1.00e-07 | 0.130 | no-effect | no-effect | 0.006 | none_required | False | -| spinosad | 1.00e-06 | 0.403 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| spinosad | 1.00e-05 | -6.704 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| sulfoxaflor | 1.00e-08 | 4.78e-04 | no-effect | no-effect | 0.006 | none_required | False | -| sulfoxaflor | 1.00e-07 | 0.005 | no-effect | no-effect | 0.006 | none_required | False | -| sulfoxaflor | 1.00e-06 | 0.046 | no-effect | no-effect | 0.006 | none_required | False | -| sulfoxaflor | 1.00e-05 | 0.339 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | -| thiamethoxam | 1.00e-08 | 7.17e-04 | no-effect | no-effect | 0.006 | none_required | False | -| thiamethoxam | 1.00e-07 | 0.007 | no-effect | no-effect | 0.006 | none_required | False | -| thiamethoxam | 1.00e-06 | 0.068 | no-effect | no-effect | 0.006 | none_required | False | -| thiamethoxam | 1.00e-05 | 0.419 | topology-dependent | topology-dependent | 0.006 | topology_without_weight_pairing | False | +| dieldrin | 1.00e-08 | 0.356 | topology-dependent | topology-dependent | 0.008 | wiring_with_weighted_transmitter_balance | False | +| dieldrin | 1.00e-07 | 2.358 | topology-dependent | topology-dependent | 0.008 | wiring_with_weighted_transmitter_balance | False | +| dieldrin | 1.00e-06 | 5.516 | topology-dependent | topology-dependent | 0.008 | wiring_with_weighted_transmitter_balance | False | +| dieldrin | 1.00e-05 | 5.944 | topology-dependent | topology-dependent | 0.008 | wiring_with_weighted_transmitter_balance | False | +| fipronil | 1.00e-08 | 0.870 | topology-dependent | topology-dependent | 0.008 | wiring_with_weighted_transmitter_balance | False | +| fipronil | 1.00e-07 | 4.574 | topology-dependent | topology-dependent | 0.008 | wiring_with_weighted_transmitter_balance | False | +| fipronil | 1.00e-06 | 5.944 | topology-dependent | topology-dependent | 0.008 | wiring_with_weighted_transmitter_balance | False | +| fipronil | 1.00e-05 | 5.944 | topology-dependent | topology-dependent | 0.008 | wiring_with_weighted_transmitter_balance | False | +| gaba | 1.00e-08 | -4.27e-06 | no-effect | no-effect | 0.008 | none_required | False | +| gaba | 1.00e-07 | -1.35e-04 | no-effect | no-effect | 0.008 | none_required | False | +| gaba | 1.00e-06 | -0.004 | no-effect | no-effect | 0.008 | none_required | False | +| gaba | 1.00e-05 | -0.122 | no-effect | no-effect | 0.008 | none_required | False | +| imidacloprid | 1.00e-08 | -0.458 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| imidacloprid | 1.00e-07 | -11.5 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| imidacloprid | 1.00e-06 | -12.2 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| imidacloprid | 1.00e-05 | -12.2 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| ivermectin | 1.00e-08 | -0.020 | no-effect | no-effect | 0.043 | none_required | False | +| ivermectin | 1.00e-07 | -0.045 | no-effect | no-effect | 0.022 | none_required | False | +| ivermectin | 1.00e-06 | -0.052 | no-effect | no-effect | 0.022 | none_required | False | +| ivermectin | 1.00e-05 | -0.053 | no-effect | no-effect | 0.022 | none_required | False | +| nicotine | 1.00e-08 | 0.221 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| nicotine | 1.00e-07 | -0.612 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| nicotine | 1.00e-06 | -11.0 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| nicotine | 1.00e-05 | -12.2 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| nitenpyram | 1.00e-08 | 0.453 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| nitenpyram | 1.00e-07 | -7.375 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| nitenpyram | 1.00e-06 | -12.2 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| nitenpyram | 1.00e-05 | -12.2 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| permethrin | 1.00e-08 | 1.058 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| permethrin | 1.00e-07 | 9.983 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | +| permethrin | 1.00e-06 | 29.4 | topology-dependent | topology-dependent | 0.008 | wiring_with_weighted_transmitter_balance | False | +| permethrin | 1.00e-05 | 54.2 | topology-dependent | topology-dependent | 0.016 | topology_without_weight_pairing | False | +| picrotoxin | 1.00e-08 | 0.038 | no-effect | no-effect | 0.008 | none_required | False | +| picrotoxin | 1.00e-07 | 0.356 | topology-dependent | topology-dependent | 0.008 | wiring_with_weighted_transmitter_balance | False | +| picrotoxin | 1.00e-06 | 2.358 | topology-dependent | topology-dependent | 0.008 | wiring_with_weighted_transmitter_balance | False | +| picrotoxin | 1.00e-05 | 5.516 | topology-dependent | topology-dependent | 0.008 | wiring_with_weighted_transmitter_balance | False | +| spinosad | 1.00e-08 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| spinosad | 1.00e-07 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| spinosad | 1.00e-06 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| spinosad | 1.00e-05 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| sulfoxaflor | 1.00e-08 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| sulfoxaflor | 1.00e-07 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| sulfoxaflor | 1.00e-06 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| sulfoxaflor | 1.00e-05 | 0.000 | no-effect | no-effect | 1.000 | none_required | False | +| thiamethoxam | 1.00e-08 | 7.17e-04 | no-effect | no-effect | 0.008 | none_required | False | +| thiamethoxam | 1.00e-07 | 0.007 | no-effect | no-effect | 0.008 | none_required | False | +| thiamethoxam | 1.00e-06 | 0.068 | no-effect | no-effect | 0.008 | none_required | False | +| thiamethoxam | 1.00e-05 | 0.419 | topology-dependent | topology-dependent | 0.008 | topology_without_weight_pairing | False | diff --git a/papers/tables/T13_ablation.csv b/papers/tables/T13_ablation.csv index 816c0cc..6393dca 100644 --- a/papers/tables/T13_ablation.csv +++ b/papers/tables/T13_ablation.csv @@ -3,7 +3,7 @@ acetamiprid,neonicotinoid,1e-08,0.12660394045839746,0.5434273849238309,0.8666947 acetylcholine,endogenous transmitter,1e-08,0.0002511255632614619,0.002181662009594021,0.0016995002106785506,0.0007607670748654982,-0.0016994287526204133,0.002181662009590468 caffeine,methylxanthine,1e-08,0.0,0.0,0.0,0.0,0.0,0.0 chlordimeform,formamidine,1e-08,0.00033322225924691765,0.0,0.0022551077086276905,0.0,-0.002254981892195751,0.0 -chlorpyrifos_oxon,organophosphate (oxon metabolite),1e-08,-0.5,47.82948490230905,3.581995353392724,41.89007154138836,-3.284963244755702,47.82948490230905 +chlorpyrifos_oxon,organophosphate (oxon metabolite),1e-08,0.0,0.0,0.0,0.0,0.0,0.0 clothianidin,neonicotinoid,1e-08,0.21109548169995912,0.28567483537513993,1.4577714989444743,0.10000529217663079,-1.4068481558545844,0.28567483537513993 ddt,organochlorine,1e-08,0.0001538224888478696,0.0024942193617985,0.0010409900876169331,0.003091907228121471,-0.0010409632768650923,0.0021729134170644926 deltamethrin,type II pyrethroid,1e-08,0.18867924528301888,3.059418881329808,1.299846281792596,4.40482935703646,-1.2592332084664735,2.665303797044139 @@ -12,19 +12,19 @@ dieldrin,cyclodiene organochlorine,1e-08,-0.09090909090909091,1.0479573712255768 fipronil,phenylpyrazole,1e-08,-0.21109548169995912,2.4334097268787467,1.4577714989444743,0.16838507486121923,-1.4068481558545844,2.4334097268787467 gaba,endogenous transmitter,1e-08,2.8284191247688173e-06,-1.3041875751795828e-05,-1.914100907640659e-05,-8.671055358178137e-07,-1.914100907640659e-05,-1.3041875751795828e-05 imidacloprid,neonicotinoid,1e-08,2.128371201276027,-0.5619540335517534,-6.16942921588497,-0.19448112604907042,-6.16942921588497,-0.561954033551757 -ivermectin,avermectin,1e-08,0.3432343234323432,0.11349383606211383,-2.2062443457865166,-0.015564152084219351,-2.2062443457865166,-0.25785015446482795 +ivermectin,avermectin,1e-08,0.3333333333333333,0.15914742451154495,-2.2062443457865166,-0.012538678824231653,-2.2062443457865166,-0.21219656601539683 nicotine,alkaloid,1e-08,0.03573258979278535,0.26632603274667943,0.2425580711801283,0.09320704732198237,-0.24111095251410664,0.2663260327466759 nitenpyram,neonicotinoid,1e-08,0.12660394045839746,0.5434273849238309,0.8666947904179922,0.19091781755964465,-0.8484771650770231,0.5434273849238309 permethrin,type I pyrethroid,1e-08,0.024390243902439025,0.3954858553914189,0.16540112475481283,0.5000584660495315,-0.1647269834625824,0.34453927132521933 picrotoxin,plant convulsant,1e-08,-0.009900990099009903,0.11413397112357693,0.06705978169748672,0.00760206859210566,-0.06694870199260894,0.11413397112357693 -spinosad,spinosyn,1e-08,0.003992015968063872,0.017219238892167255,0.013510023612548494,0.006005747612804591,-0.01350550927303118,0.017219238892167255 -sulfoxaflor,sulfoximine,1e-08,6.666222251849878e-05,0.0005795581048921861,0.000451131735961674,0.00020209322619546555,-0.00045112670062152915,0.0005795581048886334 +spinosad,spinosyn,1e-08,0.0,0.0,0.0,0.0,0.0,0.0 +sulfoxaflor,sulfoximine,1e-08,0.0,0.0,0.0,0.0,0.0,0.0 thiamethoxam,neonicotinoid,1e-08,9.999000099990002e-05,0.0008691922629786575,0.000676676937835019,0.00030309050857635356,-0.0006766656090730905,0.0008691922629751048 acetamiprid,neonicotinoid,1e-07,0.6967304549770723,-10.80994671403197,-4.5486923560284005,-3.377888997795963,-4.5486923560284005,-9.417406749555951 acetylcholine,endogenous transmitter,1e-07,0.003965285619152201,0.03393625144298085,0.026843564886990023,0.011839023579403474,-0.02682574850180952,0.03393625144298085 caffeine,methylxanthine,1e-07,0.0,0.0,0.0,0.0,0.0,0.0 chlordimeform,formamidine,1e-07,0.0033222591362126247,0.0,0.022489289137381796,0.0,-0.02247678259486463,0.0 -chlorpyrifos_oxon,organophosphate (oxon metabolite),1e-07,-0.9090909090909091,-10.80994671403197,-5.907823488940227,-4.453314692141862,-5.907823488940227,-9.417406749555951 +chlorpyrifos_oxon,organophosphate (oxon metabolite),1e-07,0.0,0.0,0.0,0.0,0.0,0.0 clothianidin,neonicotinoid,1e-07,0.8091918198317981,-10.80994671403197,-5.268790309171518,-4.773623977900712,-5.268790309171518,-9.417406749555951 ddt,organochlorine,1e-07,0.0015360983102918587,0.02490771132265479,0.010396713702393257,0.030911376259251888,-0.01039404003389599,0.021699093892983967 deltamethrin,type II pyrethroid,1e-07,0.6993006993006993,11.339104944788787,5.1727214554210015,24.021540644552797,-4.565162624979351,9.87839868834541 @@ -33,19 +33,19 @@ dieldrin,cyclodiene organochlorine,1e-07,-0.5,5.763765541740675,3.58199535339272 fipronil,phenylpyrazole,1e-07,-0.8091918198317981,9.327983855609894,6.107512111326622,0.7380224886401772,-5.268790309171518,9.327983855609894 gaba,endogenous transmitter,1e-07,8.943471981546934e-05,-0.0004123846050472224,-0.0006052350437188991,-2.741773682846116e-05,-0.0006052350437188991,-0.0004123846050472224 imidacloprid,neonicotinoid,1e-07,2.852958904733035,-10.80994671403197,-6.16942921588497,-5.680388639120631,-6.16942921588497,-9.417406749555951 -ivermectin,avermectin,1e-07,0.9242424242424243,-0.021314387211363695,-5.423255799443355,-0.0587293185688571,-5.423255799443355,-0.9496743635287164 +ivermectin,avermectin,1e-07,0.8333333333333334,0.39786856127886594,-5.423255799443355,-0.03121681365661999,-5.423255799443355,-0.5304914150384867 nicotine,alkaloid,1e-07,0.31811200018174035,-0.7536968357448401,-2.1071546515434,-0.26018164807203803,-2.1071546515434,-0.7536968357448401 nitenpyram,neonicotinoid,1e-07,0.6967304549770723,-10.80994671403197,-4.5486923560284005,-3.377888997795963,-4.5486923560284005,-9.417406749555951 permethrin,type I pyrethroid,1e-07,0.2,3.2429840142095934,1.3794979934542546,4.710059982330625,-1.3338268721244448,2.8252220248667825 picrotoxin,plant convulsant,1e-07,-0.0909090909090909,1.0479573712255768,0.6202059367630355,0.07085762022170528,-0.6108266719415321,1.0479573712255768 -spinosad,spinosyn,1e-07,0.039215686274509796,0.15714136453870964,0.13291474746747323,0.054912834542182765,-0.1324790742611972,0.15714136453870964 -sulfoxaflor,sulfoximine,1e-07,0.0006662225183211193,0.005778211444827264,0.004508838792724568,0.0020150179661975187,-0.004508335862204227,0.005778211444827264 +spinosad,spinosyn,1e-07,0.0,0.0,0.0,0.0,0.0,0.0 +sulfoxaflor,sulfoximine,1e-07,0.0,0.0,0.0,0.0,0.0,0.0 thiamethoxam,neonicotinoid,1e-07,0.0009990009990009988,0.008652866785851643,0.006761194590095165,0.0030176057607382845,-0.006760063749816858,0.008652866785851643 acetamiprid,neonicotinoid,1e-06,0.973270096840775,-10.80994671403197,-6.16942921588497,-6.17278216613422,-6.16942921588497,-9.417406749555951 acetylcholine,endogenous transmitter,1e-06,0.0593509431027676,0.39360018262419416,0.4037312084746043,0.13799221453647537,-0.3997371307719213,0.3936001826241906 caffeine,methylxanthine,1e-06,0.0,0.0,0.0,0.0,0.0,0.0 chlordimeform,formamidine,1e-06,0.03225806451612903,0.0,0.21890590735543558,0.0,-0.21772658591980498,0.0 -chlorpyrifos_oxon,organophosphate (oxon metabolite),1e-06,-0.9900990099009902,-10.80994671403197,-6.16942921588497,-5.555436983853854,-6.16942921588497,-9.417406749555951 +chlorpyrifos_oxon,organophosphate (oxon metabolite),1e-06,0.0,0.0,0.0,0.0,0.0,0.0 clothianidin,neonicotinoid,1e-06,0.9853400782049834,-10.80994671403197,-6.16942921588497,-6.17278216613422,-6.16942921588497,-9.417406749555951 ddt,organochlorine,1e-06,0.015151515151515152,0.24568060713709272,0.10266765574594316,0.30831405448478044,-0.10240751972358098,0.2140319715808161 deltamethrin,type II pyrethroid,1e-06,0.9587727708533077,15.546423845683567,7.460714474074884,37.15887655509541,-6.16942921588497,13.543729745286598 @@ -54,19 +54,19 @@ dieldrin,cyclodiene organochlorine,1e-06,-0.909090909090909,10.479573712255771,6 fipronil,phenylpyrazole,1e-06,-0.9853400782049834,10.951154529307281,7.712092780851213,0.8996076235547861,-6.16942921588497,10.951154529307281 gaba,endogenous transmitter,1e-06,0.002820449688343696,-0.013005128580712011,-0.019082583373513984,-0.0008644848379635306,-0.019082583373513984,-0.013005128580712011 imidacloprid,neonicotinoid,1e-06,2.9888577908821548,-10.80994671403197,-6.16942921588497,-6.17278216613422,-6.16942921588497,-9.417406749555951 -ivermectin,avermectin,1e-06,1.4803921568627452,-1.837425556368192,-6.16942921588497,-0.1835528137105662,-6.16942921588497,-2.9296137638003685 +ivermectin,avermectin,1e-06,0.9803921568627451,0.46808066032807716,-6.16942921588497,-0.03668098746430726,-6.16942921588497,-0.6241075471040993 nicotine,alkaloid,1e-06,0.8545051332184401,-10.80994671403197,-5.5586974000953315,-5.40399067063418,-5.5586974000953315,-9.417406749555951 nitenpyram,neonicotinoid,1e-06,0.973270096840775,-10.80994671403197,-6.16942921588497,-6.17278216613422,-6.16942921588497,-9.417406749555951 permethrin,type I pyrethroid,1e-06,0.7142857142857143,11.582085765034254,5.2975399928402975,24.820376182767554,-4.661173294012422,10.090078660238511 picrotoxin,plant convulsant,1e-06,-0.5,5.763765541740675,3.581995353392724,0.4236252131450913,-3.284963244755702,5.763765541740675 -spinosad,spinosyn,1e-06,0.3333333333333333,0.4831261101243314,1.1455623019106262,0.16958996890116396,-1.113920028244741,0.4831261101243314 -sulfoxaflor,sulfoximine,1e-06,0.006622516556291391,0.056065597446998794,0.04484211184624609,0.019564984263337415,-0.0447924161155564,0.05606559744699524 +spinosad,spinosyn,1e-06,0.0,0.0,0.0,0.0,0.0,0.0 +sulfoxaflor,sulfoximine,1e-06,0.0,0.0,0.0,0.0,0.0,0.0 thiamethoxam,neonicotinoid,1e-06,0.009900990099009901,0.08269171534919195,0.06705978169748672,0.028867082651210296,-0.06694870199260894,0.08269171534919195 acetamiprid,neonicotinoid,1e-05,0.9982701355087407,-10.80994671403197,-6.16942921588497,-6.17278216613422,-6.16942921588497,-9.417406749555951 acetylcholine,endogenous transmitter,1e-05,0.5,-4.348134991119007,-3.284963244755702,-1.4378006330560869,-3.284963244755702,-4.348134991119011 caffeine,methylxanthine,1e-05,0.0,0.0,0.0,0.0,0.0,0.0 chlordimeform,formamidine,1e-05,0.25,0.0,1.7339008885200506,0.0,-1.662233377362102,0.0 -chlorpyrifos_oxon,organophosphate (oxon metabolite),1e-05,-0.999000999000999,-10.80994671403197,-6.16942921588497,-5.549928143951719,-6.16942921588497,-9.417406749555951 +chlorpyrifos_oxon,organophosphate (oxon metabolite),1e-05,0.0,0.0,0.0,0.0,0.0,0.0 clothianidin,neonicotinoid,1e-05,0.9990621400136929,-10.80994671403197,-6.16942921588497,-6.17278216613422,-6.16942921588497,-9.417406749555951 ddt,organochlorine,1e-05,0.13333333333333336,2.161989342806395,0.9133684203732662,2.9816464493304498,-0.8931559375447362,1.8834813499111878 deltamethrin,type II pyrethroid,1e-05,0.9957184108334164,16.145494444934744,7.811271941754921,38.16064501855905,-6.16942921588497,14.065627924259609 @@ -75,11 +75,11 @@ dieldrin,cyclodiene organochlorine,1e-05,-0.99009900990099,10.951154529307281,7. fipronil,phenylpyrazole,1e-05,-0.9990621400136929,10.951154529307281,7.84334520573964,0.8996076235547861,-6.16942921588497,10.951154529307281 gaba,endogenous transmitter,1e-05,0.08209951522176571,-0.3785618854630606,-0.5519960219908775,-0.025019861031331914,-0.5519960219908775,-0.3785618854630606 imidacloprid,neonicotinoid,1e-05,2.999214992927742,-10.80994671403197,-6.16942921588497,-6.17278216613422,-6.16942921588497,-9.417406749555951 -ivermectin,avermectin,1e-05,1.9070949011068772,-3.71534018995756,-6.16942921588497,-0.29716171988116447,-6.16942921588497,-4.82714854482864 +ivermectin,avermectin,1e-05,0.9980039920159681,0.47648929494474856,-6.16942921588497,-0.03733448908736925,-6.16942921588497,-0.6353190599263314 nicotine,alkaloid,1e-05,0.9866556142826448,-10.80994671403197,-6.16942921588497,-6.17278216613422,-6.16942921588497,-9.417406749555951 nitenpyram,neonicotinoid,1e-05,0.9982701355087407,-10.80994671403197,-6.16942921588497,-6.17278216613422,-6.16942921588497,-9.417406749555951 permethrin,type I pyrethroid,1e-05,0.9615384615384616,15.591269299084587,7.486717545963149,37.24852214329439,-6.16942921588497,13.582798196474936 picrotoxin,plant convulsant,1e-05,-0.9090909090909091,10.479573712255771,6.999949583094388,0.8512734176258983,-5.907823488940227,10.479573712255771 -spinosad,spinosyn,1e-05,1.3333333333333333,-9.662522202486675,-4.355984529746964,-3.0409209448354613,-4.355984529746964,-9.417406749555951 -sulfoxaflor,sulfoximine,1e-05,0.06250000000000001,0.40763765541740327,0.4252737684436827,0.14294144232433847,-0.4208442849023353,0.40763765541740327 +spinosad,spinosyn,1e-05,0.0,0.0,0.0,0.0,0.0,0.0 +sulfoxaflor,sulfoximine,1e-05,0.0,0.0,0.0,0.0,0.0,0.0 thiamethoxam,neonicotinoid,1e-05,0.09090909090909091,0.5030899989724453,0.6202059367630355,0.1766469540699287,-0.6108266719415321,0.5030899989724418 diff --git a/papers/tables/T13_ablation.md b/papers/tables/T13_ablation.md index edcf888..80ac602 100644 --- a/papers/tables/T13_ablation.md +++ b/papers/tables/T13_ablation.md @@ -1,4 +1,4 @@ -**T13_ablation.** Ablation ladder. Each level is a prediction for the same compound-concentration cell from a model that has been denied one layer of information: receptor engagement only (dimensionless), mechanism gains on the graph's transmitter composition only (excitation index), the real cut with one generic multiplier instead of mechanism-specific gains (Hz), and the full model (Hz). The generic multiplier is direction-aware: generic in magnitude, with one bit of sign taken from the mechanism table, which is the minimum a connectome-without-pharmacology model needs to order a library containing disinhibitors. `C_topology_only_floor` is the historical depression-only rule, kept as a floor: every one of its entries is at most zero, so it cannot express disinhibition and is not a competitive baseline. Levels have different units, so they are compared by ordering, never by value. +**T13_ablation.** Ablation ladder. Each level is a prediction for the same compound-concentration cell from a model that has been denied one layer of information: receptor engagement only (dimensionless), mechanism gains on the graph's transmitter composition only (excitation index), the real cut with one generic multiplier instead of mechanism-specific gains (rate-model units), and the full model (rate-model units). The generic multiplier is direction-aware: generic in magnitude, with one bit of sign taken from the mechanism table, which is the minimum a connectome-without-pharmacology model needs to order a library containing disinhibitors. `C_topology_only_floor` is the historical depression-only rule, kept as a floor: every one of its entries is at most zero, so it cannot express disinhibition and is not a competitive baseline. Levels have different units, so they are compared by ordering, never by value. | compound | class | conc_M | A_receptor_only | B_composition_only | C_topology_only | D_full_flylab | |---|---|---|---|---|---|---| @@ -6,7 +6,7 @@ | acetylcholine | endogenous transmitter | 1.00e-08 | 2.51e-04 | 0.002 | 0.002 | 7.61e-04 | | caffeine | methylxanthine | 1.00e-08 | 0.000 | 0.000 | 0.000 | 0.000 | | chlordimeform | formamidine | 1.00e-08 | 3.33e-04 | 0.000 | 0.002 | 0.000 | -| chlorpyrifos_oxon | organophosphate (oxon metabolite) | 1.00e-08 | -0.500 | 47.8 | 3.582 | 41.9 | +| chlorpyrifos_oxon | organophosphate (oxon metabolite) | 1.00e-08 | 0.000 | 0.000 | 0.000 | 0.000 | | clothianidin | neonicotinoid | 1.00e-08 | 0.211 | 0.286 | 1.458 | 0.100 | | ddt | organochlorine | 1.00e-08 | 1.54e-04 | 0.002 | 0.001 | 0.003 | | deltamethrin | type II pyrethroid | 1.00e-08 | 0.189 | 3.059 | 1.300 | 4.405 | @@ -15,19 +15,19 @@ | fipronil | phenylpyrazole | 1.00e-08 | -0.211 | 2.433 | 1.458 | 0.168 | | gaba | endogenous transmitter | 1.00e-08 | 2.83e-06 | -1.30e-05 | -1.91e-05 | -8.67e-07 | | imidacloprid | neonicotinoid | 1.00e-08 | 2.128 | -0.562 | -6.169 | -0.194 | -| ivermectin | avermectin | 1.00e-08 | 0.343 | 0.113 | -2.206 | -0.016 | +| ivermectin | avermectin | 1.00e-08 | 0.333 | 0.159 | -2.206 | -0.013 | | nicotine | alkaloid | 1.00e-08 | 0.036 | 0.266 | 0.243 | 0.093 | | nitenpyram | neonicotinoid | 1.00e-08 | 0.127 | 0.543 | 0.867 | 0.191 | | permethrin | type I pyrethroid | 1.00e-08 | 0.024 | 0.395 | 0.165 | 0.500 | | picrotoxin | plant convulsant | 1.00e-08 | -0.010 | 0.114 | 0.067 | 0.008 | -| spinosad | spinosyn | 1.00e-08 | 0.004 | 0.017 | 0.014 | 0.006 | -| sulfoxaflor | sulfoximine | 1.00e-08 | 6.67e-05 | 5.80e-04 | 4.51e-04 | 2.02e-04 | +| spinosad | spinosyn | 1.00e-08 | 0.000 | 0.000 | 0.000 | 0.000 | +| sulfoxaflor | sulfoximine | 1.00e-08 | 0.000 | 0.000 | 0.000 | 0.000 | | thiamethoxam | neonicotinoid | 1.00e-08 | 1.00e-04 | 8.69e-04 | 6.77e-04 | 3.03e-04 | | acetamiprid | neonicotinoid | 1.00e-07 | 0.697 | -10.8 | -4.549 | -3.378 | | acetylcholine | endogenous transmitter | 1.00e-07 | 0.004 | 0.034 | 0.027 | 0.012 | | caffeine | methylxanthine | 1.00e-07 | 0.000 | 0.000 | 0.000 | 0.000 | | chlordimeform | formamidine | 1.00e-07 | 0.003 | 0.000 | 0.022 | 0.000 | -| chlorpyrifos_oxon | organophosphate (oxon metabolite) | 1.00e-07 | -0.909 | -10.8 | -5.908 | -4.453 | +| chlorpyrifos_oxon | organophosphate (oxon metabolite) | 1.00e-07 | 0.000 | 0.000 | 0.000 | 0.000 | | clothianidin | neonicotinoid | 1.00e-07 | 0.809 | -10.8 | -5.269 | -4.774 | | ddt | organochlorine | 1.00e-07 | 0.002 | 0.025 | 0.010 | 0.031 | | deltamethrin | type II pyrethroid | 1.00e-07 | 0.699 | 11.3 | 5.173 | 24.0 | @@ -36,19 +36,19 @@ | fipronil | phenylpyrazole | 1.00e-07 | -0.809 | 9.328 | 6.108 | 0.738 | | gaba | endogenous transmitter | 1.00e-07 | 8.94e-05 | -4.12e-04 | -6.05e-04 | -2.74e-05 | | imidacloprid | neonicotinoid | 1.00e-07 | 2.853 | -10.8 | -6.169 | -5.680 | -| ivermectin | avermectin | 1.00e-07 | 0.924 | -0.021 | -5.423 | -0.059 | +| ivermectin | avermectin | 1.00e-07 | 0.833 | 0.398 | -5.423 | -0.031 | | nicotine | alkaloid | 1.00e-07 | 0.318 | -0.754 | -2.107 | -0.260 | | nitenpyram | neonicotinoid | 1.00e-07 | 0.697 | -10.8 | -4.549 | -3.378 | | permethrin | type I pyrethroid | 1.00e-07 | 0.200 | 3.243 | 1.379 | 4.710 | | picrotoxin | plant convulsant | 1.00e-07 | -0.091 | 1.048 | 0.620 | 0.071 | -| spinosad | spinosyn | 1.00e-07 | 0.039 | 0.157 | 0.133 | 0.055 | -| sulfoxaflor | sulfoximine | 1.00e-07 | 6.66e-04 | 0.006 | 0.005 | 0.002 | +| spinosad | spinosyn | 1.00e-07 | 0.000 | 0.000 | 0.000 | 0.000 | +| sulfoxaflor | sulfoximine | 1.00e-07 | 0.000 | 0.000 | 0.000 | 0.000 | | thiamethoxam | neonicotinoid | 1.00e-07 | 9.99e-04 | 0.009 | 0.007 | 0.003 | | acetamiprid | neonicotinoid | 1.00e-06 | 0.973 | -10.8 | -6.169 | -6.173 | | acetylcholine | endogenous transmitter | 1.00e-06 | 0.059 | 0.394 | 0.404 | 0.138 | | caffeine | methylxanthine | 1.00e-06 | 0.000 | 0.000 | 0.000 | 0.000 | | chlordimeform | formamidine | 1.00e-06 | 0.032 | 0.000 | 0.219 | 0.000 | -| chlorpyrifos_oxon | organophosphate (oxon metabolite) | 1.00e-06 | -0.990 | -10.8 | -6.169 | -5.555 | +| chlorpyrifos_oxon | organophosphate (oxon metabolite) | 1.00e-06 | 0.000 | 0.000 | 0.000 | 0.000 | | clothianidin | neonicotinoid | 1.00e-06 | 0.985 | -10.8 | -6.169 | -6.173 | | ddt | organochlorine | 1.00e-06 | 0.015 | 0.246 | 0.103 | 0.308 | | deltamethrin | type II pyrethroid | 1.00e-06 | 0.959 | 15.5 | 7.461 | 37.2 | @@ -57,19 +57,19 @@ | fipronil | phenylpyrazole | 1.00e-06 | -0.985 | 11.0 | 7.712 | 0.900 | | gaba | endogenous transmitter | 1.00e-06 | 0.003 | -0.013 | -0.019 | -8.64e-04 | | imidacloprid | neonicotinoid | 1.00e-06 | 2.989 | -10.8 | -6.169 | -6.173 | -| ivermectin | avermectin | 1.00e-06 | 1.480 | -1.837 | -6.169 | -0.184 | +| ivermectin | avermectin | 1.00e-06 | 0.980 | 0.468 | -6.169 | -0.037 | | nicotine | alkaloid | 1.00e-06 | 0.855 | -10.8 | -5.559 | -5.404 | | nitenpyram | neonicotinoid | 1.00e-06 | 0.973 | -10.8 | -6.169 | -6.173 | | permethrin | type I pyrethroid | 1.00e-06 | 0.714 | 11.6 | 5.298 | 24.8 | | picrotoxin | plant convulsant | 1.00e-06 | -0.500 | 5.764 | 3.582 | 0.424 | -| spinosad | spinosyn | 1.00e-06 | 0.333 | 0.483 | 1.146 | 0.170 | -| sulfoxaflor | sulfoximine | 1.00e-06 | 0.007 | 0.056 | 0.045 | 0.020 | +| spinosad | spinosyn | 1.00e-06 | 0.000 | 0.000 | 0.000 | 0.000 | +| sulfoxaflor | sulfoximine | 1.00e-06 | 0.000 | 0.000 | 0.000 | 0.000 | | thiamethoxam | neonicotinoid | 1.00e-06 | 0.010 | 0.083 | 0.067 | 0.029 | | acetamiprid | neonicotinoid | 1.00e-05 | 0.998 | -10.8 | -6.169 | -6.173 | | acetylcholine | endogenous transmitter | 1.00e-05 | 0.500 | -4.348 | -3.285 | -1.438 | | caffeine | methylxanthine | 1.00e-05 | 0.000 | 0.000 | 0.000 | 0.000 | | chlordimeform | formamidine | 1.00e-05 | 0.250 | 0.000 | 1.734 | 0.000 | -| chlorpyrifos_oxon | organophosphate (oxon metabolite) | 1.00e-05 | -0.999 | -10.8 | -6.169 | -5.550 | +| chlorpyrifos_oxon | organophosphate (oxon metabolite) | 1.00e-05 | 0.000 | 0.000 | 0.000 | 0.000 | | clothianidin | neonicotinoid | 1.00e-05 | 0.999 | -10.8 | -6.169 | -6.173 | | ddt | organochlorine | 1.00e-05 | 0.133 | 2.162 | 0.913 | 2.982 | | deltamethrin | type II pyrethroid | 1.00e-05 | 0.996 | 16.1 | 7.811 | 38.2 | @@ -78,11 +78,11 @@ | fipronil | phenylpyrazole | 1.00e-05 | -0.999 | 11.0 | 7.843 | 0.900 | | gaba | endogenous transmitter | 1.00e-05 | 0.082 | -0.379 | -0.552 | -0.025 | | imidacloprid | neonicotinoid | 1.00e-05 | 2.999 | -10.8 | -6.169 | -6.173 | -| ivermectin | avermectin | 1.00e-05 | 1.907 | -3.715 | -6.169 | -0.297 | +| ivermectin | avermectin | 1.00e-05 | 0.998 | 0.476 | -6.169 | -0.037 | | nicotine | alkaloid | 1.00e-05 | 0.987 | -10.8 | -6.169 | -6.173 | | nitenpyram | neonicotinoid | 1.00e-05 | 0.998 | -10.8 | -6.169 | -6.173 | | permethrin | type I pyrethroid | 1.00e-05 | 0.962 | 15.6 | 7.487 | 37.2 | | picrotoxin | plant convulsant | 1.00e-05 | -0.909 | 10.5 | 7.000 | 0.851 | -| spinosad | spinosyn | 1.00e-05 | 1.333 | -9.663 | -4.356 | -3.041 | -| sulfoxaflor | sulfoximine | 1.00e-05 | 0.063 | 0.408 | 0.425 | 0.143 | +| spinosad | spinosyn | 1.00e-05 | 0.000 | 0.000 | 0.000 | 0.000 | +| sulfoxaflor | sulfoximine | 1.00e-05 | 0.000 | 0.000 | 0.000 | 0.000 | | thiamethoxam | neonicotinoid | 1.00e-05 | 0.091 | 0.503 | 0.620 | 0.177 | diff --git a/papers/tables/T13b_ablation_information.csv b/papers/tables/T13b_ablation_information.csv index 32d7c02..50e49f7 100644 --- a/papers/tables/T13b_ablation_information.csv +++ b/papers/tables/T13b_ablation_information.csv @@ -1,21 +1,21 @@ conc_M,level,unit,spearman_rho_vs_full,pearson_r_vs_full,residual_rms_standardised,residual_rms_hz,reproduces_full_ordering,information_added_vs_previous -1e-08,A_receptor_only,dimensionless,-0.04620893838817564,-0.289170928700701,1.6057216002163646,,False, -1e-08,B_composition_only,excitation index,0.9061889250814332,0.9976765037278486,0.06816885318312374,,True,0.9523978634696089 -1e-08,C_topology_only,Hz,0.9283856134172466,0.46968595214607956,1.0298679991668065,8.514714959614382,True,0.02219668833581334 -1e-08,D_full_flylab,Hz,1.0,1.0,0.0,0.0,True,0.07161438658275343 -1e-08,C_topology_only_floor,Hz,-0.42708342384411907,-0.3691981009986054,1.6548100199108087,10.05058983763574,False, -1e-07,A_receptor_only,dimensionless,-0.3481941132066756,-0.06440515883844936,1.459044316556868,,False, -1e-07,B_composition_only,excitation index,0.9885475201437236,0.7171532039770323,0.7521260479772893,,True,1.3367416333503992 -1e-07,C_topology_only,Hz,0.9625777447339687,0.6382780302472236,0.8505550772910317,4.59304520728783,True,-0.02596977540975487 -1e-07,D_full_flylab,Hz,1.0,1.0,0.0,0.0,True,0.03742225526603127 -1e-07,C_topology_only_floor,Hz,0.27864589238610205,0.054206195576548453,1.3753499950365011,6.6961190442005245,False, -1e-06,A_receptor_only,dimensionless,-0.4314654500371033,0.03101366199423331,1.3921108705888097,,False, -1e-06,B_composition_only,excitation index,0.9878103669773378,0.7555164920023544,0.699261765003129,,True,1.419275817014441 -1e-06,C_topology_only,Hz,0.9610814391588707,0.6537352062402837,0.8321836260822683,8.149381803321555,True,-0.02672892781846714 -1e-06,D_full_flylab,Hz,1.0,0.9999999999999998,0.0,0.0,True,0.038918560841129324 -1e-06,C_topology_only_floor,Hz,0.24159822391955538,-0.06960725750473676,1.4626053859498374,11.739863157102864,False, -1e-05,A_receptor_only,dimensionless,-0.5057464187739324,0.04378625273001715,1.3829054539410732,,False, -1e-05,B_composition_only,excitation index,0.9870768926883093,0.753923095927088,0.7015367475377355,,True,1.4928233114622418 -1e-05,C_topology_only,Hz,0.9393463731260359,0.6507608928587834,0.8357500908061174,9.72906541402147,True,-0.04773051956227348 -1e-05,D_full_flylab,Hz,1.0,1.0,0.0,0.0,True,0.06065362687396414 -1e-05,C_topology_only_floor,Hz,0.23627498954419957,-0.11764339152920802,1.4950875502987833,13.895705893886532,False, +1e-08,A_receptor_only,dimensionless,0.09527077257487362,-0.03301816905736719,1.4373713292377635,,False, +1e-08,B_composition_only,excitation index,0.8763297872340425,0.7603070679378396,0.6923769667777234,,False,0.7810590146591689 +1e-08,C_topology_only,Hz,0.923901376977036,0.2677237517741642,1.210186967559836,1.6185280637169492,True,0.04757158974299347 +1e-08,D_full_flylab,Hz,0.9999999999999999,1.0,0.0,0.0,True,0.07609862302296388 +1e-08,C_topology_only_floor,Hz,-0.3712812840144107,-0.043241827780173624,1.4444665643622032,1.9582620279112293,False, +1e-07,A_receptor_only,dimensionless,-0.5193580310505262,-0.1357909561113432,1.5071768019123326,,False, +1e-07,B_composition_only,excitation index,0.9183102371956979,0.7061936698801082,0.7665589737520419,,True,1.4376682682462243 +1e-07,C_topology_only,Hz,0.9619701619712934,0.620208776980361,0.8715402721844113,4.583571094279977,True,0.04365992477559544 +1e-07,D_full_flylab,Hz,0.9999999999999999,1.0,0.0,0.0,True,0.03802983802870652 +1e-07,C_topology_only_floor,Hz,0.1747206042420756,-0.014104889690300878,1.4241523020311424,6.6895205938088615,False, +1e-06,A_receptor_only,dimensionless,-0.5640523883085501,-0.008285426718388106,1.4200601583865298,,False, +1e-06,B_composition_only,excitation index,0.9214816983710898,0.7503365230416151,0.7066307054726468,,True,1.4855340866796398 +1e-06,C_topology_only,Hz,0.9676140362132637,0.6452548150801432,0.8423125131681909,8.150695665444447,True,0.046132337842173854 +1e-06,D_full_flylab,Hz,1.0,1.0,0.0,0.0,True,0.032385963786736305 +1e-06,C_topology_only_floor,Hz,0.16087398593797766,-0.11572300821834729,1.493802535958717,11.739358877967218,False, +1e-05,A_receptor_only,dimensionless,-0.5787938915672565,0.041567813603825624,1.3845087117069177,,False, +1e-05,B_composition_only,excitation index,0.9541778975741239,0.7503567990386449,0.7066020109812242,,True,1.5329717891413803 +1e-05,C_topology_only,Hz,0.9504884338392682,0.6362141436861005,0.8529781431125881,9.731331737956676,True,-0.0036894637348556802 +1e-05,D_full_flylab,Hz,1.0,1.0,0.0,0.0,True,0.049511566160731824 +1e-05,C_topology_only_floor,Hz,0.15091318164688006,-0.16530824269065578,1.5266356753925645,13.896884748738845,False, diff --git a/papers/tables/T13b_ablation_information.md b/papers/tables/T13b_ablation_information.md index cbe502e..01e22dd 100644 --- a/papers/tables/T13b_ablation_information.md +++ b/papers/tables/T13b_ablation_information.md @@ -2,23 +2,23 @@ | conc_M | level | unit | spearman_rho_vs_full | pearson_r_vs_full | residual_rms_standardised | residual_rms_hz | reproduces_full_ordering | information_added_vs_previous | |---|---|---|---|---|---|---|---|---| -| 1.00e-08 | A_receptor_only | dimensionless | -0.046 | -0.289 | 1.606 | -- | False | -- | -| 1.00e-08 | B_composition_only | excitation index | 0.906 | 0.998 | 0.068 | -- | True | 0.952 | -| 1.00e-08 | C_topology_only | Hz | 0.928 | 0.470 | 1.030 | 8.515 | True | 0.022 | -| 1.00e-08 | D_full_flylab | Hz | 1.000 | 1.000 | 0.000 | 0.000 | True | 0.072 | -| 1.00e-08 | C_topology_only_floor | Hz | -0.427 | -0.369 | 1.655 | 10.1 | False | -- | -| 1.00e-07 | A_receptor_only | dimensionless | -0.348 | -0.064 | 1.459 | -- | False | -- | -| 1.00e-07 | B_composition_only | excitation index | 0.989 | 0.717 | 0.752 | -- | True | 1.337 | -| 1.00e-07 | C_topology_only | Hz | 0.963 | 0.638 | 0.851 | 4.593 | True | -0.026 | -| 1.00e-07 | D_full_flylab | Hz | 1.000 | 1.000 | 0.000 | 0.000 | True | 0.037 | -| 1.00e-07 | C_topology_only_floor | Hz | 0.279 | 0.054 | 1.375 | 6.696 | False | -- | -| 1.00e-06 | A_receptor_only | dimensionless | -0.431 | 0.031 | 1.392 | -- | False | -- | -| 1.00e-06 | B_composition_only | excitation index | 0.988 | 0.756 | 0.699 | -- | True | 1.419 | -| 1.00e-06 | C_topology_only | Hz | 0.961 | 0.654 | 0.832 | 8.149 | True | -0.027 | -| 1.00e-06 | D_full_flylab | Hz | 1.000 | 1.000 | 0.000 | 0.000 | True | 0.039 | -| 1.00e-06 | C_topology_only_floor | Hz | 0.242 | -0.070 | 1.463 | 11.7 | False | -- | -| 1.00e-05 | A_receptor_only | dimensionless | -0.506 | 0.044 | 1.383 | -- | False | -- | -| 1.00e-05 | B_composition_only | excitation index | 0.987 | 0.754 | 0.702 | -- | True | 1.493 | -| 1.00e-05 | C_topology_only | Hz | 0.939 | 0.651 | 0.836 | 9.729 | True | -0.048 | -| 1.00e-05 | D_full_flylab | Hz | 1.000 | 1.000 | 0.000 | 0.000 | True | 0.061 | -| 1.00e-05 | C_topology_only_floor | Hz | 0.236 | -0.118 | 1.495 | 13.9 | False | -- | +| 1.00e-08 | A_receptor_only | dimensionless | 0.095 | -0.033 | 1.437 | -- | False | -- | +| 1.00e-08 | B_composition_only | excitation index | 0.876 | 0.760 | 0.692 | -- | False | 0.781 | +| 1.00e-08 | C_topology_only | Hz | 0.924 | 0.268 | 1.210 | 1.619 | True | 0.048 | +| 1.00e-08 | D_full_flylab | Hz | 1.000 | 1.000 | 0.000 | 0.000 | True | 0.076 | +| 1.00e-08 | C_topology_only_floor | Hz | -0.371 | -0.043 | 1.444 | 1.958 | False | -- | +| 1.00e-07 | A_receptor_only | dimensionless | -0.519 | -0.136 | 1.507 | -- | False | -- | +| 1.00e-07 | B_composition_only | excitation index | 0.918 | 0.706 | 0.767 | -- | True | 1.438 | +| 1.00e-07 | C_topology_only | Hz | 0.962 | 0.620 | 0.872 | 4.584 | True | 0.044 | +| 1.00e-07 | D_full_flylab | Hz | 1.000 | 1.000 | 0.000 | 0.000 | True | 0.038 | +| 1.00e-07 | C_topology_only_floor | Hz | 0.175 | -0.014 | 1.424 | 6.690 | False | -- | +| 1.00e-06 | A_receptor_only | dimensionless | -0.564 | -0.008 | 1.420 | -- | False | -- | +| 1.00e-06 | B_composition_only | excitation index | 0.921 | 0.750 | 0.707 | -- | True | 1.486 | +| 1.00e-06 | C_topology_only | Hz | 0.968 | 0.645 | 0.842 | 8.151 | True | 0.046 | +| 1.00e-06 | D_full_flylab | Hz | 1.000 | 1.000 | 0.000 | 0.000 | True | 0.032 | +| 1.00e-06 | C_topology_only_floor | Hz | 0.161 | -0.116 | 1.494 | 11.7 | False | -- | +| 1.00e-05 | A_receptor_only | dimensionless | -0.579 | 0.042 | 1.385 | -- | False | -- | +| 1.00e-05 | B_composition_only | excitation index | 0.954 | 0.750 | 0.707 | -- | True | 1.533 | +| 1.00e-05 | C_topology_only | Hz | 0.950 | 0.636 | 0.853 | 9.731 | True | -0.004 | +| 1.00e-05 | D_full_flylab | Hz | 1.000 | 1.000 | 0.000 | 0.000 | True | 0.050 | +| 1.00e-05 | C_topology_only_floor | Hz | 0.151 | -0.165 | 1.527 | 13.9 | False | -- | diff --git a/papers/tables/T14_conclusion_stability.csv b/papers/tables/T14_conclusion_stability.csv index 73dd9da..1c40b4e 100644 --- a/papers/tables/T14_conclusion_stability.csv +++ b/papers/tables/T14_conclusion_stability.csv @@ -1,7 +1,7 @@ conclusion,claim,readout,n_specs,n_retained,n_lost,n_undecidable,fraction_retained,fraction_retained_decidable,n_retained_uncorrected,multiplicity,n_equivalent_within_tolerance,n_indeterminate,failing_specs,undecidable_specs C1_nicotinic_suppression,"a nicotinic agonist (imidacloprid, 1 uM) suppresses network activity","subgraph mean_hz, treated < vehicle",25,15,10,0,0.6,0.6,,,,,linear@0.50x; linear@0.75x; linear@1.00x; linear@1.25x; linear@1.50x; saturating@0.50x; saturating@0.75x; saturating@1.00x; saturating@1.25x; saturating@1.50x, C5_nicotinic_buffering,the neighborhood buffers nicotinic receptor selectivity,mean (circuit SI - receptor SI) < 0 over the nicotinic set,25,18,0,7,0.72,1.0,,,,,,linear@0.50x; linear@0.75x; saturating@0.50x; saturating@0.75x; weak_biphasic@0.50x; weak_biphasic@0.75x; weak_biphasic@1.00x -C6_nav_ache_amplification,the neighborhood amplifies Nav / AChE receptor selectivity,mean (circuit SI - receptor SI) > 0 over the Nav/AChE set,25,19,6,0,0.76,0.76,,,,,flylab_biphasic@0.50x; linear@0.50x; strong_biphasic@0.50x; strong_biphasic@0.75x; strong_biphasic@1.00x; weak_biphasic@0.50x, +C6_nav_ache_amplification,the neighborhood amplifies Nav / AChE receptor selectivity,mean (circuit SI - receptor SI) > 0 over the Nav/AChE set,25,18,7,0,0.72,0.72,,,,,flylab_biphasic@0.50x; linear@0.50x; strong_biphasic@0.50x; strong_biphasic@0.75x; strong_biphasic@1.00x; weak_biphasic@0.50x; weak_biphasic@0.75x, C2_rdl_disinhibition,"an RDL antagonist (fipronil, 1 uM) disinhibits the network","subgraph mean_hz, treated > vehicle",25,25,0,0,1.0,1.0,,,,,, C3_imidacloprid_topology_not_distinguishable,imidacloprid's mean-rate effect is not distinguishable from any structure-preserving degradation of the cut (not topology-dependent),"dependence_landscape class != topology-dependent for imidacloprid: neither weight_permute nor rewire_degree_preserving reaches the empirical two-sided permutation threshold, n = 100 shuffles per specification (resolution 1/(n+1) = 0.0099), alpha = 0.05, Benjamini-Hochberg across the 100 structural tests of this run; a failure to reject, not a claim of equivalence",25,25,0,0,1.0,1.0,25,benjamini-hochberg across the 100 structural tests of this run (25 specifications x 2 compounds x 2 modes),17,8,, C4_fipronil_topology_exceeds,"fipronil's mean-rate effect is topology-dependent where imidacloprid's is not, on the same cut, shuffles and criterion","dependence_landscape class == topology-dependent for fipronil: at least one of weight_permute / rewire_degree_preserving is distinguishable from the real cut, n = 100 shuffles per specification (resolution 1/(n+1) = 0.0099), alpha = 0.05, Benjamini-Hochberg across the 100 structural tests of this run",25,25,0,0,1.0,1.0,25,benjamini-hochberg across the 100 structural tests of this run (25 specifications x 2 compounds x 2 modes),0,0,, diff --git a/papers/tables/T14_conclusion_stability.md b/papers/tables/T14_conclusion_stability.md index 83ba0ab..69f8a8e 100644 --- a/papers/tables/T14_conclusion_stability.md +++ b/papers/tables/T14_conclusion_stability.md @@ -4,7 +4,7 @@ |---|---|---|---|---|---|---|---| | C1_nicotinic_suppression | 25 | 15 | 10 | 0 | -- | -- | -- | | C5_nicotinic_buffering | 25 | 18 | 0 | 7 | -- | -- | -- | -| C6_nav_ache_amplification | 25 | 19 | 6 | 0 | -- | -- | -- | +| C6_nav_ache_amplification | 25 | 18 | 7 | 0 | -- | -- | -- | | C2_rdl_disinhibition | 25 | 25 | 0 | 0 | -- | -- | -- | | C3_imidacloprid_topology_not_distinguishable | 25 | 25 | 0 | 0 | 25 | 17 | 8 | | C4_fipronil_topology_exceeds | 25 | 25 | 0 | 0 | 25 | 0 | 0 | diff --git a/papers/tables/T14b_specification_diagnostics.csv b/papers/tables/T14b_specification_diagnostics.csv index 12b9bcd..24ecf46 100644 --- a/papers/tables/T14b_specification_diagnostics.csv +++ b/papers/tables/T14b_specification_diagnostics.csv @@ -1,26 +1,26 @@ spec,g_ach_imidacloprid,g_gaba_fipronil,mean_si_gap_nicotinic,mean_si_gap_nav_ache,runtime_s -linear@0.50x,1.1981873583955516,0.5073299608975084,,-0.08765807516937074,97.65 -linear@0.75x,1.2972810375933275,0.26099494134626244,,0.11706173732775427,108.24 -linear@1.00x,1.3963747167911034,0.05,-1.7336718357793948,0.28402260467315865,65.67 -linear@1.25x,1.495468395988879,0.05,-1.042055158102632,0.41338260436219354,52.98 -linear@1.50x,1.594562075186655,0.05,-0.7902143245129317,0.5044140826656136,50.4 -saturating@0.50x,1.1996348239458756,0.5014833892761252,,0.5881775663648283,47.01 -saturating@0.75x,1.2994522359188136,0.25222508391418785,,0.8366202724570322,31.78 -saturating@1.00x,1.3992696478917515,0.05,-1.1498771238118892,0.9999814789506893,30.58 -saturating@1.25x,1.4990870598646893,0.05,-0.44326159312185115,1.090565339024282,32.37 -saturating@1.50x,1.5989044718376273,0.05,-0.19858955890429944,1.1695331193516088,32.0 -weak_biphasic@0.50x,0.8054050681174831,0.26460619346825764,,-0.22844703018609702,32.86 -weak_biphasic@0.75x,0.7081076021762248,0.05,,0.05944540973643142,31.64 -weak_biphasic@1.00x,0.6108101362349666,0.05,,0.23213466736483301,30.04 -weak_biphasic@1.25x,0.5135126702937083,0.05,-2.4070917140311687,0.3719866053107854,33.63 -weak_biphasic@1.50x,0.41621520435244963,0.05,-1.6166781605797254,0.4751549231740453,31.86 -flylab_biphasic@0.50x,0.4126227778394148,0.5073299608975084,-1.509877039527341,-0.10139259461008687,27.8 -flylab_biphasic@0.75x,0.11893416675912216,0.26099494134626244,-1.0603058001482384,0.1123796848664973,32.2 -flylab_biphasic@1.00x,0.05,0.05,-0.8836082400406419,0.2769363642937981,28.14 -flylab_biphasic@1.25x,0.05,0.05,-0.7789639692699317,0.4083440745197833,24.09 -flylab_biphasic@1.50x,0.05,0.05,-0.7063800942351652,0.5007859243023947,30.98 -strong_biphasic@0.50x,0.05,0.05,-0.7506792736424797,-1.0079972613989858,33.02 -strong_biphasic@0.75x,0.05,0.05,-0.5804750451139211,-0.8304679927702834,25.08 -strong_biphasic@1.00x,0.05,0.05,-0.5065732492522648,-0.4888120067113579,23.01 -strong_biphasic@1.25x,0.05,0.05,-0.46514097779817754,0.18146371081839674,21.66 -strong_biphasic@1.50x,0.05,0.05,-0.424281662704153,0.31174769027230265,23.04 +linear@0.50x,1.1981873583955516,0.5073299608975084,,-0.18881851132185046,15.19 +linear@0.75x,1.2972810375933275,0.26099494134626244,,0.039916136876659736,15.73 +linear@1.00x,1.3963747167911034,0.05,-1.7336718357793934,0.19829975661541,16.04 +linear@1.25x,1.495468395988879,0.05,-1.042055158102632,0.32488425079114214,17.54 +linear@1.50x,1.594562075186655,0.05,-0.7902143245129317,0.42386510829144053,16.62 +saturating@0.50x,1.1996348239458756,0.5014833892761252,,0.503143494572302,18.41 +saturating@0.75x,1.2994522359188136,0.25222508391418785,,0.737907657713,16.91 +saturating@1.00x,1.3992696478917515,0.05,-1.1498771238118886,0.9159840212898943,18.01 +saturating@1.25x,1.4990870598646893,0.05,-0.443261593121851,1.0176185131922146,17.48 +saturating@1.50x,1.5989044718376273,0.05,-0.19858955890429944,1.0981738939060475,18.3 +weak_biphasic@0.50x,0.8054050681174831,0.26460619346825764,,-0.34557387618918234,17.81 +weak_biphasic@0.75x,0.7081076021762248,0.05,,-0.028215297448571763,18.03 +weak_biphasic@1.00x,0.6108101362349666,0.05,,0.14823672904242458,18.1 +weak_biphasic@1.25x,0.5135126702937083,0.05,-2.4070917140311687,0.27862103218893675,17.61 +weak_biphasic@1.50x,0.41621520435244963,0.05,-1.6166781605797254,0.3904351585152571,18.0 +flylab_biphasic@0.50x,0.4126227778394148,0.5073299608975084,-1.509877039527341,-0.18881851132185046,17.62 +flylab_biphasic@0.75x,0.11893416675912216,0.26099494134626244,-1.0603058001482384,0.039916136876659736,18.22 +flylab_biphasic@1.00x,0.05,0.05,-0.8836082400406419,0.19829975661541,18.62 +flylab_biphasic@1.25x,0.05,0.05,-0.7789639692699317,0.32488425079114214,13.5 +flylab_biphasic@1.50x,0.05,0.05,-0.7063800942351652,0.42386510829144053,12.98 +strong_biphasic@0.50x,0.05,0.05,-0.7506792736424797,-1.1351040473249046,11.73 +strong_biphasic@0.75x,0.05,0.05,-0.5804750451139211,-0.930483490265734,10.95 +strong_biphasic@1.00x,0.05,0.05,-0.5065732492522648,-0.792266550308343,11.25 +strong_biphasic@1.25x,0.05,0.05,-0.46514097779817754,0.03231442321988309,10.74 +strong_biphasic@1.50x,0.05,0.05,-0.424281662704153,0.20834439518749426,10.86 diff --git a/papers/tables/T14b_specification_diagnostics.md b/papers/tables/T14b_specification_diagnostics.md index 6341176..c9c4b17 100644 --- a/papers/tables/T14b_specification_diagnostics.md +++ b/papers/tables/T14b_specification_diagnostics.md @@ -2,28 +2,28 @@ | spec | g_ach_imidacloprid | g_gaba_fipronil | mean_si_gap_nicotinic | mean_si_gap_nav_ache | runtime_s | |---|---|---|---|---|---| -| linear@0.50x | 1.198 | 0.507 | -- | -0.088 | 97.7 | -| linear@0.75x | 1.297 | 0.261 | -- | 0.117 | 108 | -| linear@1.00x | 1.396 | 0.050 | -1.734 | 0.284 | 65.7 | -| linear@1.25x | 1.495 | 0.050 | -1.042 | 0.413 | 53.0 | -| linear@1.50x | 1.595 | 0.050 | -0.790 | 0.504 | 50.4 | -| saturating@0.50x | 1.200 | 0.501 | -- | 0.588 | 47.0 | -| saturating@0.75x | 1.299 | 0.252 | -- | 0.837 | 31.8 | -| saturating@1.00x | 1.399 | 0.050 | -1.150 | 1.000 | 30.6 | -| saturating@1.25x | 1.499 | 0.050 | -0.443 | 1.091 | 32.4 | -| saturating@1.50x | 1.599 | 0.050 | -0.199 | 1.170 | 32.0 | -| weak_biphasic@0.50x | 0.805 | 0.265 | -- | -0.228 | 32.9 | -| weak_biphasic@0.75x | 0.708 | 0.050 | -- | 0.059 | 31.6 | -| weak_biphasic@1.00x | 0.611 | 0.050 | -- | 0.232 | 30.0 | -| weak_biphasic@1.25x | 0.514 | 0.050 | -2.407 | 0.372 | 33.6 | -| weak_biphasic@1.50x | 0.416 | 0.050 | -1.617 | 0.475 | 31.9 | -| flylab_biphasic@0.50x | 0.413 | 0.507 | -1.510 | -0.101 | 27.8 | -| flylab_biphasic@0.75x | 0.119 | 0.261 | -1.060 | 0.112 | 32.2 | -| flylab_biphasic@1.00x | 0.050 | 0.050 | -0.884 | 0.277 | 28.1 | -| flylab_biphasic@1.25x | 0.050 | 0.050 | -0.779 | 0.408 | 24.1 | -| flylab_biphasic@1.50x | 0.050 | 0.050 | -0.706 | 0.501 | 31.0 | -| strong_biphasic@0.50x | 0.050 | 0.050 | -0.751 | -1.008 | 33.0 | -| strong_biphasic@0.75x | 0.050 | 0.050 | -0.580 | -0.830 | 25.1 | -| strong_biphasic@1.00x | 0.050 | 0.050 | -0.507 | -0.489 | 23.0 | -| strong_biphasic@1.25x | 0.050 | 0.050 | -0.465 | 0.181 | 21.7 | -| strong_biphasic@1.50x | 0.050 | 0.050 | -0.424 | 0.312 | 23.0 | +| linear@0.50x | 1.198 | 0.507 | -- | -0.189 | 15.2 | +| linear@0.75x | 1.297 | 0.261 | -- | 0.040 | 15.7 | +| linear@1.00x | 1.396 | 0.050 | -1.734 | 0.198 | 16.0 | +| linear@1.25x | 1.495 | 0.050 | -1.042 | 0.325 | 17.5 | +| linear@1.50x | 1.595 | 0.050 | -0.790 | 0.424 | 16.6 | +| saturating@0.50x | 1.200 | 0.501 | -- | 0.503 | 18.4 | +| saturating@0.75x | 1.299 | 0.252 | -- | 0.738 | 16.9 | +| saturating@1.00x | 1.399 | 0.050 | -1.150 | 0.916 | 18.0 | +| saturating@1.25x | 1.499 | 0.050 | -0.443 | 1.018 | 17.5 | +| saturating@1.50x | 1.599 | 0.050 | -0.199 | 1.098 | 18.3 | +| weak_biphasic@0.50x | 0.805 | 0.265 | -- | -0.346 | 17.8 | +| weak_biphasic@0.75x | 0.708 | 0.050 | -- | -0.028 | 18.0 | +| weak_biphasic@1.00x | 0.611 | 0.050 | -- | 0.148 | 18.1 | +| weak_biphasic@1.25x | 0.514 | 0.050 | -2.407 | 0.279 | 17.6 | +| weak_biphasic@1.50x | 0.416 | 0.050 | -1.617 | 0.390 | 18.0 | +| flylab_biphasic@0.50x | 0.413 | 0.507 | -1.510 | -0.189 | 17.6 | +| flylab_biphasic@0.75x | 0.119 | 0.261 | -1.060 | 0.040 | 18.2 | +| flylab_biphasic@1.00x | 0.050 | 0.050 | -0.884 | 0.198 | 18.6 | +| flylab_biphasic@1.25x | 0.050 | 0.050 | -0.779 | 0.325 | 13.5 | +| flylab_biphasic@1.50x | 0.050 | 0.050 | -0.706 | 0.424 | 13.0 | +| strong_biphasic@0.50x | 0.050 | 0.050 | -0.751 | -1.135 | 11.7 | +| strong_biphasic@0.75x | 0.050 | 0.050 | -0.580 | -0.930 | 10.9 | +| strong_biphasic@1.00x | 0.050 | 0.050 | -0.507 | -0.792 | 11.2 | +| strong_biphasic@1.25x | 0.050 | 0.050 | -0.465 | 0.032 | 10.7 | +| strong_biphasic@1.50x | 0.050 | 0.050 | -0.424 | 0.208 | 10.9 | diff --git a/papers/tables/T15_threshold_grid.csv b/papers/tables/T15_threshold_grid.csv index af92d99..c1677d8 100644 --- a/papers/tables/T15_threshold_grid.csv +++ b/papers/tables/T15_threshold_grid.csv @@ -1,10 +1,10 @@ circuit_frac,vert_limit,n_amplify,n_buffer,n_unscored,mean_gap_nicotinic,verdict_nicotinic,mean_gap_nav_ache,verdict_nav_ache -0.25,0.1,4,5,1,-0.9288320271187136,buffer,0.22689219345980904,amplify -0.25,0.2,4,5,1,-0.5883889262777295,buffer,0.5790747115711716,amplify -0.25,0.3,4,5,1,-0.362108493778807,buffer,0.8131579176045398,amplify -0.4,0.1,1,8,1,-1.111252611511875,buffer,0.037714001833427324,amplify -0.4,0.2,4,5,1,-0.7708095106708911,buffer,0.3898965199447899,amplify -0.4,0.3,4,5,1,-0.5445290781719685,buffer,0.6239797259781581,amplify -0.5,0.1,1,8,1,-1.2240513408816258,buffer,-0.0752461538175645,buffer -0.5,0.2,4,5,1,-0.8836082400406419,buffer,0.2769363642937981,amplify -0.5,0.3,4,5,1,-0.6573278075417192,buffer,0.5110195703271663,amplify +0.25,0.1,2,5,3,-0.9288320271187136,buffer,0.1661649733360372,amplify +0.25,0.2,2,5,3,-0.5883889262777298,buffer,0.5183474914473997,amplify +0.25,0.3,2,5,3,-0.36210849377880716,buffer,0.7524306974807679,amplify +0.4,0.1,0,7,3,-1.111252611511875,buffer,-0.041654034083071445,buffer +0.4,0.2,2,5,3,-0.7708095106708911,buffer,0.31052848402829103,amplify +0.4,0.3,2,5,3,-0.5445290781719685,buffer,0.5446116900616592,amplify +0.5,0.1,0,7,3,-1.2240513408816258,buffer,-0.15388276149595248,buffer +0.5,0.2,2,5,3,-0.8836082400406419,buffer,0.19829975661541,amplify +0.5,0.3,2,5,3,-0.6573278075417192,buffer,0.4323829626487782,amplify diff --git a/papers/tables/T15_threshold_grid.md b/papers/tables/T15_threshold_grid.md index 84ce67f..d400fb4 100644 --- a/papers/tables/T15_threshold_grid.md +++ b/papers/tables/T15_threshold_grid.md @@ -2,12 +2,12 @@ | circuit_frac | vert_limit | n_amplify | n_buffer | n_unscored | mean_gap_nicotinic | verdict_nicotinic | mean_gap_nav_ache | verdict_nav_ache | |---|---|---|---|---|---|---|---|---| -| 0.250 | 0.100 | 4 | 5 | 1 | -0.929 | buffer | 0.227 | amplify | -| 0.250 | 0.200 | 4 | 5 | 1 | -0.588 | buffer | 0.579 | amplify | -| 0.250 | 0.300 | 4 | 5 | 1 | -0.362 | buffer | 0.813 | amplify | -| 0.400 | 0.100 | 1 | 8 | 1 | -1.111 | buffer | 0.038 | amplify | -| 0.400 | 0.200 | 4 | 5 | 1 | -0.771 | buffer | 0.390 | amplify | -| 0.400 | 0.300 | 4 | 5 | 1 | -0.545 | buffer | 0.624 | amplify | -| 0.500 | 0.100 | 1 | 8 | 1 | -1.224 | buffer | -0.075 | buffer | -| 0.500 | 0.200 | 4 | 5 | 1 | -0.884 | buffer | 0.277 | amplify | -| 0.500 | 0.300 | 4 | 5 | 1 | -0.657 | buffer | 0.511 | amplify | +| 0.250 | 0.100 | 2 | 5 | 3 | -0.929 | buffer | 0.166 | amplify | +| 0.250 | 0.200 | 2 | 5 | 3 | -0.588 | buffer | 0.518 | amplify | +| 0.250 | 0.300 | 2 | 5 | 3 | -0.362 | buffer | 0.752 | amplify | +| 0.400 | 0.100 | 0 | 7 | 3 | -1.111 | buffer | -0.042 | buffer | +| 0.400 | 0.200 | 2 | 5 | 3 | -0.771 | buffer | 0.311 | amplify | +| 0.400 | 0.300 | 2 | 5 | 3 | -0.545 | buffer | 0.545 | amplify | +| 0.500 | 0.100 | 0 | 7 | 3 | -1.224 | buffer | -0.154 | buffer | +| 0.500 | 0.200 | 2 | 5 | 3 | -0.884 | buffer | 0.198 | amplify | +| 0.500 | 0.300 | 2 | 5 | 3 | -0.657 | buffer | 0.432 | amplify | diff --git a/papers/tables/T16_uncertainty_budget.csv b/papers/tables/T16_uncertainty_budget.csv index 0c8fecb..12b81e7 100644 --- a/papers/tables/T16_uncertainty_budget.csv +++ b/papers/tables/T16_uncertainty_budget.csv @@ -1,11 +1,11 @@ source,kind,share_of_variance_S1,share_with_interactions_ST,variance_removed_hz2,ci_low,ci_high gain_transform,categorical,0.4096497698742678,0.6264806479027828,1.290081356802269,0.31256086807181116,0.5001624331362007 -weight_threshold,continuous,0.29357557224259245,0.5424303725151057,0.9245370079884832,0.1987612666853245,0.39130919604916053 -drive,continuous,0.018555443445925905,0.14121674202113985,0.05843535970772074,-0.0716609947161836,0.11220171357678527 -transmitter,continuous,0.006078670560203601,0.023642782102253644,0.019143131866687924,-0.05210207681142646,0.06965580923052817 -expression,continuous,-0.005809124285986339,0.00012806053442048645,-0.01829426864562489,-0.058783087902346,0.044256230041565946 -potency,continuous,-0.005989859757710479,0.0006784194893530697,-0.018863446220546365,-0.06115133451219928,0.045925733350771346 -lif_seed,seed,-0.006741189415684353,0.0,-0.021229556141375117,-0.06185506580751919,0.043437093161159034 +weight_threshold,continuous,0.29357557224259245,0.5424303725151058,0.9245370079884832,0.1987612666853245,0.39130919604916053 +drive,continuous,0.018555443445925905,0.14121674202113985,0.05843535970772074,-0.0716609947161836,0.11220171357678516 +transmitter,continuous,0.006078670560203601,0.023642782102253644,0.019143131866687924,-0.05210207681142619,0.06965580923052817 +expression,continuous,-0.00580912428598648,0.0001280605344204865,-0.018294268645625333,-0.058783087902346,0.04425623004156553 +potency,continuous,-0.005989859757710479,0.0006784194893530694,-0.018863446220546365,-0.06115133451219928,0.045925733350771346 +lif_seed,seed,-0.006741189415684353,0.0,-0.021229556141375117,-0.06185506580751933,0.04343709316115891 hill_n,continuous,-0.007946568350353221,0.0007525711738826337,-0.025025571679174075,-0.06210277198208717,0.04046709008223071 -gain_coef,continuous,-0.04565580230843003,0.023822223893828185,-0.14378062364354172,-0.10739018244206512,0.02765398020264121 -interactions (higher order),residual,0.3442830879951747,,1.0842266393099944,, +gain_coef,continuous,-0.04565580230843003,0.023822223893828185,-0.14378062364354172,-0.10739018244206512,0.027653980202641208 +interactions (higher order),residual,0.3442830879951748,,1.0842266393099949,, diff --git a/papers/tables/T17_value_of_information.csv b/papers/tables/T17_value_of_information.csv index 56772d8..e267765 100644 --- a/papers/tables/T17_value_of_information.csv +++ b/papers/tables/T17_value_of_information.csv @@ -1,10 +1,10 @@ -rank,factor,state,voi_fraction_of_var,voi_fraction_raw,clipped_from_negative,voi_var_hz2,voi_upper_bound_var_hz2,sd_reduction_hz,experiment,resolves,cost,blocking_gate -1,gain_transform,resolved,0.4096497698742678,0.4096497698742678,False,1.290081356802269,1.9729316691788883,0.4111009848265976,calibration of receptor engagement against synaptic gain,the functional form of the engagement -> gain transformation,high,the single fitted parameter the project allows -2,weight_threshold,resolved,0.29357557224259245,0.29357557224259245,False,0.9245370079884832,1.7082380179533077,0.2830665210214218,no experiment: a reconstruction-confidence analysis,how many weak edges belong in the cut,none (compute only), -3,drive,null control,0.018555443445925905,0.018555443445925905,False,0.05843535970772074,0.4447240046192244,0.016541402539729466,in vivo baseline firing rates of the driven cell classes,the external drive rate the network is simulated at,medium, -4,transmitter,null control,0.006078670560203601,0.006078670560203601,False,0.019143131866687924,0.07445655937367515,0.005401847210167565,immunostaining or ground-truth transmitter labels,the probability that a predicted transmitter -- hence an edge's sign -- is wrong,medium, -5,expression,null control,0.0,-0.005809124285986339,True,0.0,0.0004032920805709512,0.0,FISH / scRNA-seq for receptor expression in the named cell class,"which cells actually carry the receptor the drug acts on; FlyLab's expression coverage is 0.21 and adult motor-neuron receptor expression -- MN9, the main readout -- has no source at all",medium,v0.6 gate 1 (expression coverage) -6,potency,null control,0.0,-0.005989859757710479,True,0.0,0.0021364990283635147,0.0,electrophysiological dose-response on the stated receptor,the sourced potency value and its parameter type (EC50 vs IC50 vs Kd),medium, -7,lif_seed,null control,0.0,-0.006741189415684353,True,0.0,0.0,0.0,none: increase the number of simulated replicates,nothing about biology,none (compute only), -8,hill_n,null control,0.0,-0.007946568350353221,True,0.0,0.002370019740010527,0.0,"the same dose-response, read for slope rather than midpoint",the cooperativity of the engagement curve,low (rides along with the potency ladder), -9,gain_coef,null control,0.0,-0.04565580230843003,True,0.0,0.07502166285222114,0.0,"the same calibration, read for amplitude rather than shape",the size of the gain-rule coefficients,medium,the single fitted parameter the project allows +rank,factor,state,voi_fraction_of_var,voi_fraction_raw,clipped_from_negative,voi_var_hz2,voi_upper_bound_var_hz2,sd_reduction_hz,experiment,resolves,cost,blocking_gate +1,gain_transform,resolved,0.4096497698742678,0.4096497698742678,False,1.290081356802269,1.9729316691788883,0.4111009848265976,calibration of receptor engagement against synaptic gain,the functional form of the engagement -> gain transformation,high,the single fitted parameter the project allows +2,weight_threshold,resolved,0.29357557224259245,0.29357557224259245,False,0.9245370079884832,1.7082380179533079,0.2830665210214218,no experiment: a reconstruction-confidence analysis,how many weak edges belong in the cut,none (compute only), +3,drive,null control,0.018555443445925905,0.018555443445925905,False,0.05843535970772074,0.4447240046192244,0.016541402539729466,in vivo baseline firing rates of the driven cell classes,the external drive rate the network is simulated at,medium, +4,transmitter,null control,0.006078670560203601,0.006078670560203601,False,0.019143131866687924,0.07445655937367515,0.005401847210167565,immunostaining or ground-truth transmitter labels,the probability that a predicted transmitter -- hence an edge's sign -- is wrong,medium, +5,expression,null control,0.0,-0.00580912428598648,True,0.0,0.0004032920805709514,0.0,FISH / scRNA-seq for receptor expression in the named cell class,"which cells actually carry the receptor the drug acts on; only 0.205 of graph-node × receptor-key pairs currently have a coarse class-level annotation, and MN9-specific receptor expression remains unresolved",medium,v0.7 gate 7 (receptor-expression evidence) +6,potency,null control,0.0,-0.005989859757710479,True,0.0,0.0021364990283635143,0.0,electrophysiological dose-response on the stated receptor,the sourced potency value and its parameter type (EC50 vs IC50 vs Kd),medium, +7,lif_seed,null control,0.0,-0.006741189415684353,True,0.0,0.0,0.0,none: increase the number of simulated replicates,nothing about biology,none (compute only), +8,hill_n,null control,0.0,-0.007946568350353221,True,0.0,0.002370019740010527,0.0,"the same dose-response, read for slope rather than midpoint",the cooperativity of the engagement curve,low (rides along with the potency ladder), +9,gain_coef,null control,0.0,-0.04565580230843003,True,0.0,0.07502166285222114,0.0,"the same calibration, read for amplitude rather than shape",the size of the gain-rule coefficients,medium,the single fitted parameter the project allows diff --git a/papers/tables/T17_value_of_information.md b/papers/tables/T17_value_of_information.md index 6ac91f5..99be0f3 100644 --- a/papers/tables/T17_value_of_information.md +++ b/papers/tables/T17_value_of_information.md @@ -1,4 +1,4 @@ -**T17_value_of_information.** Value of information. VOI_j = S_j x Var(Y) is the model variance that would disappear if assumption j were resolved exactly while everything else stayed as uncertain as it is; the upper bound uses the total-order index and is what resolving j *last* would buy. A value of information cannot be negative, so a negative first-order estimate is clipped to zero for the decision value and kept unclipped in `voi_fraction_raw`. `state` says whether the sample could resolve the factor at all: an unresolved factor carries no ranking claim, which is a statement about the sample size rather than about the factor. This is variance of a model output under assumed input ranges, not an expected gain in accuracy about a living fly. +**T17_value_of_information.** Author-defined variance-priority heuristic. H_j = max(0, S_j) × Var(Y) ranks variance associated with factor j under assumed input ranges; it is not a formal expected value of information. The upper-bound column uses the total-order index. The heuristic cannot be negative, so a negative first-order estimate is clipped to zero for the decision value and kept unclipped in `voi_fraction_raw`. `state` says whether the sample could resolve the factor at all: an unresolved factor carries no ranking claim, which is a statement about the sample size rather than about the factor. This uses squared rate-model units, despite legacy column names ending in hz2; it is not an expected gain in accuracy about a living fly. | rank | factor | state | voi_fraction_of_var | voi_var_hz2 | experiment | cost | |---|---|---|---|---|---|---| diff --git a/papers/tables/T18_claim_provenance.csv b/papers/tables/T18_claim_provenance.csv index 20140c4..033ffe3 100644 --- a/papers/tables/T18_claim_provenance.csv +++ b/papers/tables/T18_claim_provenance.csv @@ -1,10 +1,10 @@ -step,label,classification,contributes,assumptions,unknowns -result,COMPUTED,PREDICTION,"subgraph readout for imidacloprid at 1.00e-06 M. It is a simulation output, not a measurement from a fly.",Every number downstream of this point is model output., -parameter_source,LITERATURE-DERIVED,LITERATURE,"4 of 6 receptor rows carry a sourced, typed potency or affinity parameter; the rest are class placeholders and are reported as not modelled, never as zero.",,no sourced value at vertebrate_GABA_A; no sourced value at insect_RDL -engagement_transformation,COMPUTED,MODEL-DERIVED,"Engagement is the Hill expression C^n / (value^n + C^n) applied to that typed parameter. Only a Kd/Ki row measured on this receptor in this species (evidence distance E0) yields fractional receptor occupancy; a Kd/Ki from another species or a related preparation yields a binding-derived engagement proxy, and an EC50/IC50 row yields normalised functional engagement (a proxy in turn when it was transferred here). None of the proxies is the same quantity as an occupancy.",A single-site Hill relation with the library's n is assumed at every receptor.; Assay conditions behind an EC50/IC50 are not carried through the transformation., -mechanism_rule,MODEL-ASSUMPTION,MODEL-ASSUMPTION,"Engagement becomes a synaptic gain through a rule FlyLab asserts. The shape of that rule -- monotone, saturating or rise-then-fall -- has never been measured for these receptors; it is the project's central modelling choice, not a result.","The occupancy-to-gain transformation is asserted, not fitted to any animal data.; The same rule is applied to every synapse of a transmitter class.", -expression_assumption,MODEL-ASSUMPTION,MODEL-ASSUMPTION,"The gain is applied uniformly to every cell of a transmitter class, because per-cell receptor expression is largely unmapped (20.5% of cells mapped). A cell that does not express the receptor is patched exactly like one that does.",Uniform receptor expression across the patched transmitter class.,receptor expression in MN9 and in every other adult motor neuron -malecns_edges,OBSERVED,LITERATURE,"The wiring is measured: a hops-limited neighbourhood of the public MaleCNS v1.0 reconstruction, with a synapse-count floor. This is the one layer of the chain that is a direct observation of the system being modelled.","The cut is a neighbourhood with a weight floor, not the whole CNS.", -transmitter_predictions,MODEL-ASSUMPTION,PREDICTION,"Edge signs come from MaleCNS *predicted* consensus transmitters, not from immunostaining. A wrong prediction flips a synapse from excitatory to inhibitory, and the drug patch follows the label, not the biology.",Predicted transmitter labels are taken as ground truth for the sign of every edge.,ground-truth transmitter identity for the cells in this cut -engine,MODEL-ASSUMPTION,MODEL-ASSUMPTION,"The rate engine turns patched weights into activity. Its time constants, drive and settling are teaching choices; they are not fitted to any recording.","Sensory drive amplitude is set by hand, not measured.; No fitted parameter separates the engine from the data.", -readout,COMPUTED,PREDICTION,"The reported rates are the engine's output on the patched graph. They are predictions about a simulation, and carry no animal units.",,how any of these numbers maps to a behavioural or electrophysiological measurement +step,label,classification,contributes,assumptions,unknowns +result,COMPUTED,PREDICTION,"subgraph readout for imidacloprid at 1.00e-06 M. It is a simulation output, not a measurement from a fly.",Every number downstream of this point is model output., +parameter_source,LITERATURE-DERIVED,LITERATURE,"4 of 6 receptor rows carry a sourced, typed potency or affinity parameter; the rest are class placeholders and are reported as not modelled, never as zero.",,no sourced value at vertebrate_GABA_A; no sourced value at insect_RDL +engagement_transformation,COMPUTED,MODEL-DERIVED,"Engagement is the Hill expression C^n / (value^n + C^n) applied to that typed parameter. Only a Kd/Ki row measured on this receptor in this species (evidence distance E0) yields fractional receptor occupancy; a Kd/Ki from another species or a related preparation yields a binding-derived engagement proxy, and an EC50/IC50 row yields normalised functional engagement (a proxy in turn when it was transferred here). None of the proxies is the same quantity as an occupancy.",A single-site Hill relation with the library's n is assumed at every receptor.; Assay conditions behind an EC50/IC50 are not carried through the transformation., +mechanism_rule,MODEL-ASSUMPTION,MODEL-ASSUMPTION,"Engagement becomes a synaptic gain through a rule FlyLab asserts. The shape of that rule -- monotone, saturating or rise-then-fall -- has never been measured for these receptors; it is the project's central modelling choice, not a result.","The occupancy-to-gain transformation is asserted, not fitted to any animal data.; The same rule is applied to every synapse of a transmitter class.", +expression_assumption,MODEL-ASSUMPTION,MODEL-ASSUMPTION,"The gain is applied uniformly to every cell of a transmitter class. Coarse cross-atlas class-level annotation is available for 20.5% of graph-node × receptor-key pairs; this is annotation availability, not a per-cell expression measurement. A cell that does not express the receptor is patched exactly like one that does.",Uniform receptor expression across the patched transmitter class.,receptor expression in MN9 and in every other adult motor neuron +malecns_edges,OBSERVED,LITERATURE,"The wiring is measured: a hops-limited neighbourhood of the public MaleCNS v1.0 reconstruction, with a synapse-count floor. This is the one layer of the chain that is a direct observation of the system being modelled.","The cut is a neighbourhood with a weight floor, not the whole CNS.", +transmitter_predictions,MODEL-ASSUMPTION,PREDICTION,"Edge signs come from MaleCNS *predicted* consensus transmitters, not from immunostaining. A wrong prediction flips a synapse from excitatory to inhibitory, and the drug patch follows the label, not the biology.",Predicted transmitter labels are taken as ground truth for the sign of every edge.,ground-truth transmitter identity for the cells in this cut +engine,MODEL-ASSUMPTION,MODEL-ASSUMPTION,"The rate engine turns patched weights into activity. Its time constants, drive and settling are teaching choices; they are not fitted to any recording.","Sensory drive amplitude is set by hand, not measured.; No fitted parameter separates the engine from the data.", +readout,COMPUTED,PREDICTION,"The reported rates are the engine's output on the patched graph. They are predictions about a simulation, and carry no animal units.",,how any of these numbers maps to a behavioural or electrophysiological measurement diff --git a/papers/tables/T18_claim_provenance.md b/papers/tables/T18_claim_provenance.md index b007f55..e7162cd 100644 --- a/papers/tables/T18_claim_provenance.md +++ b/papers/tables/T18_claim_provenance.md @@ -6,7 +6,7 @@ | parameter_source | LITERATURE-DERIVED | LITERATURE | 4 of 6 receptor rows carry a sourced, typed potency or affinity parameter; the rest are class placeholders and are reported as not modelled, never as zero. | | engagement_transformation | COMPUTED | MODEL-DERIVED | Engagement is the Hill expression C^n / (value^n + C^n) applied to that typed parameter. Only a Kd/Ki row measured on this receptor in this species (evidence distance E0) yields fractional receptor occupancy; a Kd/Ki from another species or a related preparation yields a binding-derived engagement proxy, and an EC50/IC50 row yields normalised functional engagement (a proxy in turn when it was transferred here). None of the proxies is the same quantity as an occupancy. | | mechanism_rule | MODEL-ASSUMPTION | MODEL-ASSUMPTION | Engagement becomes a synaptic gain through a rule FlyLab asserts. The shape of that rule -- monotone, saturating or rise-then-fall -- has never been measured for these receptors; it is the project's central modelling choice, not a result. | -| expression_assumption | MODEL-ASSUMPTION | MODEL-ASSUMPTION | The gain is applied uniformly to every cell of a transmitter class, because per-cell receptor expression is largely unmapped (20.5% of cells mapped). A cell that does not express the receptor is patched exactly like one that does. | +| expression_assumption | MODEL-ASSUMPTION | MODEL-ASSUMPTION | The gain is applied uniformly to every cell of a transmitter class. Coarse cross-atlas class-level annotation is available for 20.5% of graph-node × receptor-key pairs; this is annotation availability, not a per-cell expression measurement. A cell that does not express the receptor is patched exactly like one that does. | | malecns_edges | OBSERVED | LITERATURE | The wiring is measured: a hops-limited neighbourhood of the public MaleCNS v1.0 reconstruction, with a synapse-count floor. This is the one layer of the chain that is a direct observation of the system being modelled. | | transmitter_predictions | MODEL-ASSUMPTION | PREDICTION | Edge signs come from MaleCNS *predicted* consensus transmitters, not from immunostaining. A wrong prediction flips a synapse from excitatory to inhibitory, and the drug patch follows the label, not the biology. | | engine | MODEL-ASSUMPTION | MODEL-ASSUMPTION | The rate engine turns patched weights into activity. Its time constants, drive and settling are teaching choices; they are not fitted to any recording. | diff --git a/papers/tables/T1_library.csv b/papers/tables/T1_library.csv index fdadca4..af5b032 100644 --- a/papers/tables/T1_library.csv +++ b/papers/tables/T1_library.csv @@ -19,7 +19,7 @@ chlordimeform,Chlordimeform,formamidine,6164-98-3,insect_nAChR,insect,unknown,,, chlordimeform,Chlordimeform,formamidine,6164-98-3,insect_RDL,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder chlordimeform,Chlordimeform,formamidine,6164-98-3,vertebrate_nAChR_a4b2,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,"class-order placeholder: mammalian action is at alpha2-adrenoceptors, which are not in this receptor set" chlordimeform,Chlordimeform,formamidine,6164-98-3,vertebrate_GABA_A,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder -chlorpyrifos_oxon,Chlorpyrifos-oxon,organophosphate (oxon metabolite),5598-15-2,insect_AChE,insect,IC50,1e-08,1e-08,exact_compound_exact_receptor_exact_species,Drosophila melanogaster AChE (Ace),1.0,inhibitor,1.0,literature_order,"Fournier and Mutero 1994, Comp Biochem Physiol C 108:19 (Drosophila AChE is inhibited by oxon organophosphates in the low-nM order and is more sensitive than mammalian AChE)." +chlorpyrifos_oxon,Chlorpyrifos-oxon,organophosphate (oxon metabolite),5598-15-2,insect_AChE,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,"Placeholder pending source verification: Fournier and Mutero 1994 is a resistance review, not a primary Drosophila chlorpyrifos-oxon 10 nM IC50 assay. The previous number is withheld pending a primary enzyme assay with exposure time." chlorpyrifos_oxon,Chlorpyrifos-oxon,organophosphate (oxon metabolite),5598-15-2,vertebrate_AChE,vertebrate,IC50,3e-07,3e-07,exact_compound_exact_receptor_other_species,Homo sapiens erythrocyte AChE in donor blood,1.0,inhibitor,1.0,literature_order,"Archives of Toxicology 2020, interspecies AChE inhibition study in electric eel and human donor blood (chlorpyrifos-oxon IC50 about 0.3 uM in human blood)." chlorpyrifos_oxon,Chlorpyrifos-oxon,organophosphate (oxon metabolite),5598-15-2,insect_nAChR,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,"class-order placeholder: the oxon acts on the enzyme, not directly on the receptor" chlorpyrifos_oxon,Chlorpyrifos-oxon,organophosphate (oxon metabolite),5598-15-2,vertebrate_nAChR_a4b2,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder @@ -50,20 +50,20 @@ dieldrin,Dieldrin,cyclodiene organochlorine,60-57-1,insect_nAChR,insect,unknown, dieldrin,Dieldrin,cyclodiene organochlorine,60-57-1,vertebrate_nAChR_a4b2,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder fipronil,Fipronil,phenylpyrazole,120068-37-3,insect_nAChR,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder fipronil,Fipronil,phenylpyrazole,120068-37-3,vertebrate_nAChR_a4b2,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder -fipronil,Fipronil,phenylpyrazole,120068-37-3,vertebrate_GABA_A,vertebrate,IC50,1.1e-06,1.1e-06,exact_compound_exact_receptor_other_species,Homo sapiens recombinant GABA-A ([3H]EBOB displacement),1.0,antagonist,1.0,literature_order,"Ratra and Casida 2001, Toxicol Lett 122:215 ([3H]EBOB displacement, human recombinant GABA-A): vertebrate average IC50 for fipronil is 1103 nM = 1.1e-6 M. CORRECTED in v0.5: FlyLab v0.4 carried 1.0e-05 M while citing this same paper, so the library contradicted its own citation; the value is now the cited one. SUBUNIT CAVEAT: the human beta3 HOMOpentamer is far more sensitive, 0.5-2.4 nM (Ratra, Kamita and Casida 2001, Toxicol Appl Pharmacol 172:233), so the 'vertebrate-safe' reading of this row applies to native heteromeric receptors only, not to beta3 homomers." -fipronil,Fipronil,phenylpyrazole,120068-37-3,insect_RDL,insect,IC50,3e-08,3e-08,exact_compound_exact_receptor_exact_species,Drosophila melanogaster,1.2,antagonist,1.0,literature_order,"Hosie et al. 1995, Br J Pharmacol 115:909 (fipronil block of dieldrin-sensitive Drosophila GABA receptors, tens-of-nM order). SPREAD NOTE: the literature spans ~1500-fold for this one row - 2.4-12 nM on house-fly head membranes ([3H]EBOB displacement; Cole et al. 1993, Pestic Biochem Physiol 46:47 and Ratra and Casida 2001, Toxicol Lett 122:215) versus 0.3-3 uM on the predominant in vivo splice/edit variant of Drosophila RDL in Xenopus oocytes (Lees et al. 2014, PLoS ONE 9:e97468). Library keeps the potent end (unchanged from FlyLab v0.4); it is a defensible midpoint-to-potent choice, not a measured constant." +fipronil,Fipronil,phenylpyrazole,120068-37-3,vertebrate_GABA_A,vertebrate,IC50,1.1e-06,1.1e-06,exact_compound_exact_receptor_other_species,native GABA-related chloride-channel preparations from multiple vertebrate species ([3H]EBOB displacement),1.0,antagonist,1.0,literature_order,"Hainzl, Cole and Casida 1998, Chem Res Toxicol 11:1529-1535 (DOI 10.1021/tx980157t; PMID 9860498): mean fipronil IC50 1103 nM across human, dog, mouse, chicken, quail and salmon preparations by [3H]EBOB displacement. This is a multi-species binding-site proxy, not a human recombinant functional GABA-A potency." +fipronil,Fipronil,phenylpyrazole,120068-37-3,insect_RDL,insect,IC50,3e-08,3e-08,exact_compound_related_receptor,insect GABA-related chloride-channel preparations; Drosophila RDL variant response differs by assay,1.2,antagonist,1.0,literature_order,"The 30 nM model value is an assay-transfer proxy, not a measured constant for the predominant Drosophila RDLbd I/V variant. Hainzl et al. 1998 (DOI 10.1021/tx980157t) reports 3-12 nM insect [3H]EBOB displacement; Lees et al. 2014 (DOI 10.1371/journal.pone.0097468) reports about 0.63 uM functional block on RDLbd I/V. Model sensitivity to this range is unresolved." gaba,GABA,endogenous transmitter,56-12-2,insect_RDL,insect,EC50,5e-05,5e-05,exact_compound_exact_receptor_exact_species,Drosophila melanogaster RDLbd in Xenopus oocytes,1.5,agonist,1.0,literature_order,"Lees et al. 2014, PLoS ONE 9:e97468 (RDLbd and RDLbd I/V in Xenopus oocytes; GABA pEC50 4.2-4.3, i.e. about 50 uM)." gaba,GABA,endogenous transmitter,56-12-2,vertebrate_GABA_A,vertebrate,EC50,1e-05,1e-05,exact_compound_exact_receptor_other_species,mammalian alpha1beta2gamma2 GABA-A,1.5,agonist,1.0,literature_order,"Sigel and Steinmann 2012, J Biol Chem 287:40224 (GABA-A receptors; GABA EC50 in the 5-30 uM range for alpha1beta2gamma2)." gaba,GABA,endogenous transmitter,56-12-2,insect_nAChR,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder gaba,GABA,endogenous transmitter,56-12-2,vertebrate_nAChR_a4b2,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder -imidacloprid,Imidacloprid,neonicotinoid,138261-41-3,insect_nAChR,insect,EC50,2e-08,2e-08,exact_compound_related_receptor,Ctenocephalides felis alpha1/alpha2 + Gallus gallus beta2 hybrid,1.2,agonist,1.0,literature_order,"Dederer et al. 2011, Insect Biochem Mol Biol 41:51 (Drosophila/Ctenocephalides alpha1-beta1 nAChR; neonicotinoid agonist potency in the 10-100 nM order). Value kept unchanged from FlyLab v0.4. SOURCE DESCRIPTION CORRECTED in v0.6 (schema v3): Dederer et al. expressed cat-flea Cfalpha1/Cfalpha2 with the CHICKEN beta2 subunit, i.e. a hybrid receptor - not an 'alpha1-beta1' receptor as the v0.5 source string said. The value is unchanged; only the description of what was measured is corrected. It is an agonist EC50 on a hybrid receptor, so this row is typed EC50 / exact_compound_related_receptor (evidence distance E2) and drives functional_engagement_proxy, never binding occupancy. The subunit-resolved beta1 row is insect_nAChR_beta1." +imidacloprid,Imidacloprid,neonicotinoid,138261-41-3,insect_nAChR,insect,EC50,2e-08,2e-08,exact_compound_related_receptor,Ctenocephalides felis alpha1/alpha2 + Gallus gallus beta2 hybrid,1.2,agonist,1.0,literature_order,"Dederer H, Werr M, Ilg T (2011) Insect Biochem Mol Biol 41(1):51-61. DOI 10.1016/j.ibmb.2010.09.012; PMID 20933086. Imidacloprid EC50 was 0.02 uM on cat-flea Cfalpha1/chicken Gg-beta2 hybrid receptors. The same paper also studied Drosophila alpha subunits with chicken beta2, but the 0.02 uM value is not a Drosophila receptor measurement. This unchanged library value is therefore an E2 related-receptor functional-engagement proxy, not binding occupancy." imidacloprid,Imidacloprid,neonicotinoid,138261-41-3,vertebrate_nAChR_a4b2,vertebrate,EC50,1e-05,1e-05,class_extrapolation,mammalian (review-level contrast),1.0,agonist,0.3,literature_order,"Tomizawa and Casida 2005, Annu Rev Pharmacol Toxicol 45:247 (imidacloprid selectivity for insect over vertebrate nAChR is >1000-fold; vertebrate potency in the micromolar order). Order-of-magnitude selectivity contrast, kept from FlyLab v0.4." imidacloprid,Imidacloprid,neonicotinoid,138261-41-3,vertebrate_GABA_A,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder: no neonicotinoid action at GABA-A assumed imidacloprid,Imidacloprid,neonicotinoid,138261-41-3,insect_RDL,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder: no neonicotinoid action at RDL assumed imidacloprid,Imidacloprid,neonicotinoid,138261-41-3,insect_nAChR_native_dmel,insect,Kd,2e-09,2e-09,exact_compound_exact_receptor_exact_species,Drosophila melanogaster (native head membranes; nAChR population not resolved to subunits),1.0,agonist,1.0,literature_order,"Tomizawa, Latli and Casida 1996, J Neurochem 67(4):1669-1676, doi:10.1046/j.1471-4159.1996.67041669.x, PMID 8858952 (""Head membranes of Drosophila melanogaster and Musca domestica have a single high-affinity binding site for [3H]IMI with KD values of 1-2 nM and Bmax values of 560-850 fmol/mg of protein""). This is a radioligand saturation Kd measured on the MODELLED organism, which is why it is the library's only binding_occupancy row. Two honest limits are recorded rather than smoothed over: (a) the publication states the 1-2 nM range for Drosophila and Musca jointly and does not resolve the Drosophila-only value, so FlyLab takes the conservative (weakest-affinity) end, 2.0e-9 M, and the tier stays literature_order; (b) the preparation is native head membranes and resolves NO subunit - the companion photoaffinity study (Tomizawa and Casida 1997, Neurosci Lett 237:61, PMID 9453215, doi:10.1016/s0304-3940(97)00811-2) labels a single 66 kDa ligand-binding polypeptide of the Drosophila nAChR without assigning it to a named Dalpha/Dbeta subunit - so the row sits on the preparation-resolved key insect_nAChR_native_dmel and makes no subunit claim. It does not replace the Dederer hybrid-receptor EC50 on the aggregate insect_nAChR key, which is a functional potency and stays where it was." imidacloprid,Imidacloprid,neonicotinoid,138261-41-3,insect_nAChR_beta1,insect,Kd,8.3e-11,8.3e-11,exact_compound_exact_receptor_other_species,"Myzus persicae (native beta1-containing nAChR, susceptible clone 4106A)",1.0,agonist,1.0,literature_order,"Bass et al. 2011, BMC Neuroscience 12:51, doi:10.1186/1471-2202-12-51 ([3H]-imidacloprid saturation binding on whole-body membranes: the susceptible 4106A clone has two sites, Kd 0.083 +/- 0.0085 nM (Bmax 100 fmol/mg) and Kd 1.7 +/- 0.65 nM (Bmax 298 fmol/mg); the R81T mutation in the nAChR beta1 subunit abolishes the high-affinity site, leaving a single Kd 4.14 +/- 0.39 nM site in the resistant FRC clone). FlyLab uses the high-affinity Kd = 8.3e-11 M. This is a genuine equilibrium dissociation constant from a radioligand binding assay, but it was measured in an APHID, not in Drosophila. Under the v0.6.1 rule a binding constant licenses physical binding_occupancy only with relation exact_compound_exact_receptor_exact_species, so this row sits at evidence distance E1 and drives binding_engagement_proxy: the same Hill number, reported as an engagement proxy derived from a binding constant measured in another species, and carrying a provenance warning that names the gap. The transfer of an aphid beta1 affinity to the Drosophila Dbeta1 receptor is an assumption, not a measurement." ivermectin,Ivermectin,avermectin,70288-86-7,insect_GluCl,insect,EC50,2e-08,2e-08,exact_compound_related_receptor,invertebrate GluCl (Drosophila genetics plus heterologous invertebrate GluCl),1.0,agonist,1.0,literature_order,"Kane et al. 2000, PNAS 97:13949 (drug-resistant Drosophila show GluCl is the avermectin target); invertebrate GluCl activation is reported in the 0.1-30 nM order." -ivermectin,Ivermectin,avermectin,70288-86-7,insect_RDL,insect,EC50,1e-06,1e-06,exact_compound_exact_receptor_exact_species,Drosophila melanogaster RDL (predominant in vivo variant) in Xenopus oocytes,1.0,positive_modulator,0.5,literature_order,"Lees et al. 2014, PLoS ONE 9:e97468 (ivermectin activates and potentiates the predominant in vivo RDL variant at sub-micromolar to micromolar)." +ivermectin,Ivermectin,avermectin,70288-86-7,insect_RDL,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,"Placeholder pending source verification: Lees et al. 2014 (DOI 10.1371/journal.pone.0097468) reports ivermectin inhibition of GABA-evoked RDLbd I/V currents (pIC50 6.33), with no current from 10 uM ivermectin alone. The former positive-modulator EC50 contradicted that assay; an inhibitory model mapping has not yet been validated." ivermectin,Ivermectin,avermectin,70288-86-7,vertebrate_GlyR,vertebrate,EC50,1e-06,1e-06,exact_compound_exact_receptor_other_species,rat/human alpha1 glycine receptor,1.0,positive_modulator,0.8,literature_order,"Shan et al. 2001, J Biol Chem 276:12556 (ivermectin is an unconventional agonist of the alpha1 glycine receptor, sub-micromolar to micromolar)." ivermectin,Ivermectin,avermectin,70288-86-7,vertebrate_GABA_A,vertebrate,EC50,1e-05,1e-05,exact_compound_exact_receptor_other_species,Mus musculus GABA-A,1.0,positive_modulator,0.5,literature_order,"Dawson et al. 2000, J Pharmacol Exp Ther 295:1051 (avermectin effects in mice are mediated through GABA-A; micromolar potentiation)." ivermectin,Ivermectin,avermectin,70288-86-7,insect_nAChR,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder @@ -76,22 +76,22 @@ nitenpyram,Nitenpyram,neonicotinoid,150824-47-8,vertebrate_nAChR_a4b2,vertebrate nitenpyram,Nitenpyram,neonicotinoid,150824-47-8,vertebrate_GABA_A,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder nitenpyram,Nitenpyram,neonicotinoid,150824-47-8,insect_RDL,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder permethrin,Permethrin,type I pyrethroid,52645-53-1,insect_Nav,insect,EC50,4e-07,4e-07,exact_compound_exact_receptor_exact_species,Drosophila melanogaster para in Xenopus oocytes,1.0,positive_modulator,1.0,literature_order,"Burton et al. 2011, Insect Biochem Mol Biol 41:723 (Drosophila para tail-current EC50 0.40 uM for permethrin)." -permethrin,Permethrin,type I pyrethroid,52645-53-1,vertebrate_Nav1_x,vertebrate,EC50,4e-05,4e-05,class_extrapolation,Rattus norvegicus brain type IIA sodium channels (fold-difference statement),1.0,positive_modulator,0.3,literature_order,"Smith and Soderlund 1998, Neurotoxicology 19:823 (permethrin is more than 100-fold less potent on rat brain type IIA sodium channels than on insect para)." +permethrin,Permethrin,type I pyrethroid,52645-53-1,vertebrate_Nav1_x,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,"Placeholder pending source verification: Smith and Soderlund 1998 (PMID 9863771) studied cypermethrin, not permethrin. It cannot support the former 40 uM permethrin parameter; a matched mammalian permethrin assay is required." permethrin,Permethrin,type I pyrethroid,52645-53-1,insect_nAChR,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder permethrin,Permethrin,type I pyrethroid,52645-53-1,insect_RDL,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder permethrin,Permethrin,type I pyrethroid,52645-53-1,vertebrate_GABA_A,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder picrotoxin,Picrotoxin,plant convulsant,124-87-8,insect_RDL,insect,IC50,1e-06,1e-06,exact_compound_exact_receptor_exact_species,Drosophila melanogaster,1.0,antagonist,1.0,literature_order,"Hosie et al. 1995, Br J Pharmacol 115:909 (picrotoxinin block of Drosophila GABA receptors, micromolar order); RDL oocyte IC50 varies up to about 17-fold with expression conditions (PLoS ONE 2021, RDL expression-condition study)." picrotoxin,Picrotoxin,plant convulsant,124-87-8,vertebrate_GABA_A,vertebrate,IC50,1e-06,1e-06,exact_compound_exact_receptor_other_species,mammalian GABA-A,1.0,antagonist,1.0,literature_order,"Ratra and Casida 2001, Toxicol Lett 122:215 (picrotoxinin blocks mammalian GABA-A at about 1 uM, i.e. no useful insect selectivity)." -picrotoxin,Picrotoxin,plant convulsant,124-87-8,vertebrate_GlyR,vertebrate,IC50,1e-05,1e-05,class_extrapolation,mammalian homomeric glycine receptor,1.0,antagonist,1.0,literature_order,"Ratra and Casida 2001, Toxicol Lett 122:215 and standard glycine receptor pharmacology (picrotoxin blocks homomeric GlyR in the micromolar order). Order-of-magnitude value." +picrotoxin,Picrotoxin,plant convulsant,124-87-8,vertebrate_GlyR,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,Placeholder pending source verification: The previously cited Ratra and Casida GABA-receptor paper does not assay GlyR. A direct glycine-receptor picrotoxin preparation and potency are required before this row can be quantified. picrotoxin,Picrotoxin,plant convulsant,124-87-8,insect_nAChR,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder picrotoxin,Picrotoxin,plant convulsant,124-87-8,vertebrate_nAChR_a4b2,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder -spinosad,Spinosad (spinosyn A),spinosyn,131929-60-7,insect_nAChR,insect,EC50,5e-06,5e-06,exact_compound_related_receptor,Drosophila melanogaster Dalpha6 / heterologous insect alpha6-like receptors,1.0,positive_modulator,1.0,literature_order,"Perry et al. 2015, Insect Biochem Mol Biol 64:106 (Drosophila Dalpha6 is required for spinosad action in vivo); heterologous alpha6 nAChR studies report allosteric activation in the low-micromolar order (EC50 about 5 uM). AGGREGATE ROW (v0.6): the measurement behind this number is alpha6-specific and is recorded subunit-resolved in insect_nAChR_alpha6. The same value is repeated here on the documented aggregate key (no new number was invented) so the legacy g_ach gain rule still fires; read as an aggregate it overstates the receptor population the value applies to." +spinosad,Spinosad (spinosyn A),spinosyn,131929-60-7,insect_nAChR,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,Placeholder pending source verification: Perry et al. 2015 (DOI 10.1016/j.ibmb.2015.01.017; PMID 25747008) supports Dalpha6 target involvement but not a 5 uM receptor EC50. Martelli et al. 2022 (PMID 35191376) reports low-dose antagonism of Dalpha6-mediated cholinergic response. No quantitative receptor potency or direction is admitted pending an assay-matched source. spinosad,Spinosad (spinosyn A),spinosyn,131929-60-7,vertebrate_nAChR_a4b2,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder: no mammalian nAChR potency is reported for spinosyns at insecticidal concentrations. spinosad,Spinosad (spinosyn A),spinosyn,131929-60-7,vertebrate_nAChR_a7,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder spinosad,Spinosad (spinosyn A),spinosyn,131929-60-7,insect_RDL,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder spinosad,Spinosad (spinosyn A),spinosyn,131929-60-7,vertebrate_GABA_A,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder -spinosad,Spinosad (spinosyn A),spinosyn,131929-60-7,insect_nAChR_alpha6,insect,EC50,5e-06,5e-06,exact_compound_related_receptor,Drosophila melanogaster Dalpha6 (in vivo requirement) / heterologous insect alpha6-like receptors (the exact preparation behind the 5 uM order is not pinned),1.0,positive_modulator,1.0,literature_order,"Perry et al. 2015, Insect Biochem Mol Biol 64:106 (Drosophila Dalpha6 is required for spinosad action in vivo; Dalpha6 knockouts are highly resistant); heterologous alpha6 nAChR studies report allosteric activation in the low-micromolar order (EC50 about 5 uM). SUBUNIT-RESOLVED ROW (v0.6): this is the assay the spinosad number actually came from. The identical value also appears on the aggregate insect_nAChR key so the legacy gain rules keep working; no new number was invented for the split." -sulfoxaflor,Sulfoxaflor,sulfoximine,946578-00-3,insect_nAChR,insect,EC50,0.00015,0.00015,exact_compound_related_receptor,Periplaneta americana DUM neurons (alpha-bungarotoxin-insensitive nAChR),1.0,agonist,1.0,literature_order,"Watson et al. 2011, Pestic Biochem Physiol 101:32 (sulfoxaflor fully activates alpha-bungarotoxin-insensitive nAChRs of cockroach DUM neurons, EC50 about 154 uM); Sparks et al. 2013, Pestic Biochem Physiol 107:1 (mode of action review)." +spinosad,Spinosad (spinosyn A),spinosyn,131929-60-7,insect_nAChR_alpha6,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,"Placeholder pending source verification: Dalpha6 is a spinosad target (Perry et al. 2015, DOI 10.1016/j.ibmb.2015.01.017), but that study does not measure the previously entered 5 uM receptor EC50; Martelli et al. 2022 (PMID 35191376) reports low-dose antagonism. This row is intentionally not modelled until potency and effect direction are sourced." +sulfoxaflor,Sulfoxaflor,sulfoximine,946578-00-3,insect_nAChR,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,"Placeholder pending source verification: Watson et al. 2011, Insect Biochem Mol Biol 41:432-439 (DOI 10.1016/j.ibmb.2011.01.009; PMID 21296156) supports Dalpha2/chicken beta2 receptor activity but does not verify the former 154 uM value in the checked record. A table-verified potency is required." sulfoxaflor,Sulfoxaflor,sulfoximine,946578-00-3,vertebrate_nAChR_a4b2,vertebrate,EC50,0.001,0.001,class_extrapolation,mammalian (review-level contrast),1.0,agonist,0.2,literature_order,"Sparks et al. 2013, Pestic Biochem Physiol 107:1 (low mammalian nAChR activity reported for sulfoximines). Order-of-magnitude contrast, not a measured EC50." sulfoxaflor,Sulfoxaflor,sulfoximine,946578-00-3,vertebrate_nAChR_a7,vertebrate,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder sulfoxaflor,Sulfoxaflor,sulfoximine,946578-00-3,insect_RDL,insect,unknown,,,unsupported,,1.0,none,0.0,class_placeholder,class-order placeholder diff --git a/papers/tables/T1_library.md b/papers/tables/T1_library.md index f7bf5ff..c67cab2 100644 --- a/papers/tables/T1_library.md +++ b/papers/tables/T1_library.md @@ -22,7 +22,7 @@ | chlordimeform | formamidine | insect_RDL | unknown | -- | none | class_placeholder | | chlordimeform | formamidine | vertebrate_nAChR_a4b2 | unknown | -- | none | class_placeholder | | chlordimeform | formamidine | vertebrate_GABA_A | unknown | -- | none | class_placeholder | -| chlorpyrifos_oxon | organophosphate (oxon metabolite) | insect_AChE | IC50 | 1.00e-08 | inhibitor | literature_order | +| chlorpyrifos_oxon | organophosphate (oxon metabolite) | insect_AChE | unknown | -- | none | class_placeholder | | chlorpyrifos_oxon | organophosphate (oxon metabolite) | vertebrate_AChE | IC50 | 3.00e-07 | inhibitor | literature_order | | chlorpyrifos_oxon | organophosphate (oxon metabolite) | insect_nAChR | unknown | -- | none | class_placeholder | | chlorpyrifos_oxon | organophosphate (oxon metabolite) | vertebrate_nAChR_a4b2 | unknown | -- | none | class_placeholder | @@ -66,7 +66,7 @@ | imidacloprid | neonicotinoid | insect_nAChR_native_dmel | Kd | 2.00e-09 | agonist | literature_order | | imidacloprid | neonicotinoid | insect_nAChR_beta1 | Kd | 8.30e-11 | agonist | literature_order | | ivermectin | avermectin | insect_GluCl | EC50 | 2.00e-08 | agonist | literature_order | -| ivermectin | avermectin | insect_RDL | EC50 | 1.00e-06 | positive_modulator | literature_order | +| ivermectin | avermectin | insect_RDL | unknown | -- | none | class_placeholder | | ivermectin | avermectin | vertebrate_GlyR | EC50 | 1.00e-06 | positive_modulator | literature_order | | ivermectin | avermectin | vertebrate_GABA_A | EC50 | 1.00e-05 | positive_modulator | literature_order | | ivermectin | avermectin | insect_nAChR | unknown | -- | none | class_placeholder | @@ -79,22 +79,22 @@ | nitenpyram | neonicotinoid | vertebrate_GABA_A | unknown | -- | none | class_placeholder | | nitenpyram | neonicotinoid | insect_RDL | unknown | -- | none | class_placeholder | | permethrin | type I pyrethroid | insect_Nav | EC50 | 4.00e-07 | positive_modulator | literature_order | -| permethrin | type I pyrethroid | vertebrate_Nav1_x | EC50 | 4.00e-05 | positive_modulator | literature_order | +| permethrin | type I pyrethroid | vertebrate_Nav1_x | unknown | -- | none | class_placeholder | | permethrin | type I pyrethroid | insect_nAChR | unknown | -- | none | class_placeholder | | permethrin | type I pyrethroid | insect_RDL | unknown | -- | none | class_placeholder | | permethrin | type I pyrethroid | vertebrate_GABA_A | unknown | -- | none | class_placeholder | | picrotoxin | plant convulsant | insect_RDL | IC50 | 1.00e-06 | antagonist | literature_order | | picrotoxin | plant convulsant | vertebrate_GABA_A | IC50 | 1.00e-06 | antagonist | literature_order | -| picrotoxin | plant convulsant | vertebrate_GlyR | IC50 | 1.00e-05 | antagonist | literature_order | +| picrotoxin | plant convulsant | vertebrate_GlyR | unknown | -- | none | class_placeholder | | picrotoxin | plant convulsant | insect_nAChR | unknown | -- | none | class_placeholder | | picrotoxin | plant convulsant | vertebrate_nAChR_a4b2 | unknown | -- | none | class_placeholder | -| spinosad | spinosyn | insect_nAChR | EC50 | 5.00e-06 | positive_modulator | literature_order | +| spinosad | spinosyn | insect_nAChR | unknown | -- | none | class_placeholder | | spinosad | spinosyn | vertebrate_nAChR_a4b2 | unknown | -- | none | class_placeholder | | spinosad | spinosyn | vertebrate_nAChR_a7 | unknown | -- | none | class_placeholder | | spinosad | spinosyn | insect_RDL | unknown | -- | none | class_placeholder | | spinosad | spinosyn | vertebrate_GABA_A | unknown | -- | none | class_placeholder | -| spinosad | spinosyn | insect_nAChR_alpha6 | EC50 | 5.00e-06 | positive_modulator | literature_order | -| sulfoxaflor | sulfoximine | insect_nAChR | EC50 | 1.50e-04 | agonist | literature_order | +| spinosad | spinosyn | insect_nAChR_alpha6 | unknown | -- | none | class_placeholder | +| sulfoxaflor | sulfoximine | insect_nAChR | unknown | -- | none | class_placeholder | | sulfoxaflor | sulfoximine | vertebrate_nAChR_a4b2 | EC50 | 0.001 | agonist | literature_order | | sulfoxaflor | sulfoximine | vertebrate_nAChR_a7 | unknown | -- | none | class_placeholder | | sulfoxaflor | sulfoximine | insect_RDL | unknown | -- | none | class_placeholder | diff --git a/papers/tables/T22_instrument_validation.md b/papers/tables/T22_instrument_validation.md index bcd66e4..7f562f1 100644 --- a/papers/tables/T22_instrument_validation.md +++ b/papers/tables/T22_instrument_validation.md @@ -1,4 +1,4 @@ -**T22_instrument_validation.** Ground-truth recovery and power for the dependence ladder. A recurrent cholinergic cycle of known strength is planted in a synthetic cut of the same size, density and transmitter composition as the `named` cut, and a gain patch that collapses `g_ach` removes an amplification that exists only while the cycle is intact -- so the drug *contrast*, not merely the rate, depends on the wiring. `loop_strength = 0` plants nothing and is the negative control, whose detection rate is the empirical false-positive rate. Nothing here touches the connectome or the compound library: the experiment is about the instrument. +**T22_instrument_validation.** Limited planted-cycle recovery check for the dependence ladder. A recurrent cholinergic cycle of known strength is planted in a synthetic 250-node cut with about 1500 background edges. Its transmitter mix was chosen close to the `named` cut, but size and connectivity were not matched. A gain patch that collapses `g_ach` removes an amplification that exists only while the cycle is intact, so the drug *contrast* depends on the wiring by construction. `loop_strength = 0` plants nothing and is the negative control. The small sample gives an observed control count, not a precise false-positive rate or power estimate for the paper's endpoint. Nothing here touches the connectome or compound library. | experiment | loop_strength | n | replicates | detection_rate | class | |---|---|---|---|---|---| diff --git a/papers/tables/T24_composition_reference.csv b/papers/tables/T24_composition_reference.csv index a074236..78aa003 100644 --- a/papers/tables/T24_composition_reference.csv +++ b/papers/tables/T24_composition_reference.csv @@ -1,6 +1,6 @@ condition,spearman_rho,n_compounds,draws,note -observed (library gain vectors),0.9878103669773378,21,,what the paper used to quote on its own -"library, floor-saturated compounds dropped",0.9892086330935252,15,,compounds on the 0.05 gain floor removed (the tie artefact) +observed (library gain vectors),0.9214816983710898,21,,what the paper used to quote on its own +"library, floor-saturated compounds dropped",0.8263565891472867,16,,compounds on the 0.05 gain floor removed (the tie artefact) matched reference: one gain moved per pseudo-compound,0.8467627880513988,21,50,"matched reference: one gain moved per pseudo-compound, which is the shape every shipped mechanism rule produces. This is the distribution the observed value should be read against." dense reference: every gain moved per pseudo-compound,0.807555470491887,21,50,"arbitrary gain vectors, every gain moved at once: the structural part of the correlation, with no pharmacology in it" -compound labels shuffled,0.9878103669773378,21,1,provably identical to the observed value: both levels are functions of the same gain vector +compound labels shuffled,0.9214816983710898,21,1,provably identical to the observed value: both levels are functions of the same gain vector diff --git a/papers/tables/T24_composition_reference.md b/papers/tables/T24_composition_reference.md index c6b7e7c..426e904 100644 --- a/papers/tables/T24_composition_reference.md +++ b/papers/tables/T24_composition_reference.md @@ -1,9 +1,9 @@ -**T24_composition_reference.** Reference distribution for the composition-versus-full rank correlation. The composition level and the full model are not independent models: both are functions of the same gain vector, so the correlation has a large structural floor and a bare value near 0.99 is not interpretable. The matched reference draws pseudo-compounds with the shape the shipped mechanism rules produce (one receptor, one transmitter, one gain moved) and no pharmacology at all; shuffling the compound labels leaves the observed value unchanged, which is a proof rather than a coincidence. +**T24_composition_reference.** Reference distribution for the composition-versus-full rank correlation. The composition level and the full model are not independent models: both are functions of the same gain vector, so the correlation has a large structural floor and the observed value must be compared with a matched reference. The latter draws pseudo-compounds with the shape the shipped mechanism rules produce (one receptor, one transmitter, one gain moved) and no pharmacology at all; shuffling the compound labels leaves the observed value unchanged, which is a consequence of moving the paired B and D values together. | condition | spearman_rho | n_compounds | draws | |---|---|---|---| -| observed (library gain vectors) | 0.988 | 21 | -- | -| library, floor-saturated compounds dropped | 0.989 | 15 | -- | +| observed (library gain vectors) | 0.921 | 21 | -- | +| library, floor-saturated compounds dropped | 0.826 | 16 | -- | | matched reference: one gain moved per pseudo-compound | 0.847 | 21 | 50 | | dense reference: every gain moved per pseudo-compound | 0.808 | 21 | 50 | -| compound labels shuffled | 0.988 | 21 | 1 | +| compound labels shuffled | 0.921 | 21 | 1 | diff --git a/papers/tables/T25_engine_normalisation.csv b/papers/tables/T25_engine_normalisation.csv index 2b52d73..809b8dc 100644 --- a/papers/tables/T25_engine_normalisation.csv +++ b/papers/tables/T25_engine_normalisation.csv @@ -1,15 +1,15 @@ conc_M,normalisation,spearman_rho_b_vs_d,pearson_r_b_vs_d,reproduces_full_ordering -1e-06,row_abs,0.9878103669773378,0.7555164920023544,True -1e-06,none,0.9056122378744635,0.9353196924012067,True -1e-06,degree,0.6662469964157954,0.8274849003853563,False -1e-08,row_abs,0.9061889250814332,0.9976765037278486,True -1e-08,none,0.9214681776227494,0.6527717367860683,True -1e-08,degree,-0.37198697068403913,0.7682476172083534,False -1e-08,glutamate sign: wholens (B excitatory),0.9061889250814332,0.9976765037278486, -1e-08,glutamate sign: rate_engine (B inhibitory),0.9635179153094463,0.9972395310551259, -1e-07,glutamate sign: wholens (B excitatory),0.9885475201437236,0.7171532039770322, -1e-07,glutamate sign: rate_engine (B inhibitory),0.9872364491621005,0.70046340255922, -1e-06,glutamate sign: wholens (B excitatory),0.9878103669773378,0.7555164920023544, -1e-06,glutamate sign: rate_engine (B inhibitory),0.9838459106550208,0.730164248651969, -1e-05,glutamate sign: wholens (B excitatory),0.9870768926883093,0.7539230959270876, -1e-05,glutamate sign: rate_engine (B inhibitory),0.9786946657808683,0.7233006778568019, +1e-06,row_abs,0.9214816983710898,0.7503365230416149,True +1e-06,none,0.9348816268044955,0.9326721668623552,True +1e-06,degree,0.5510525399252045,0.8054253759393977,False +1e-08,row_abs,0.8763297872340425,0.7603070679378398,False +1e-08,none,0.9091349346404287,0.8500558824362258,True +1e-08,degree,-0.5832082533976854,-0.35236054136583606,False +1e-08,glutamate sign: wholens (B excitatory),0.8763297872340425,0.7603070679378398, +1e-08,glutamate sign: rate_engine (B inhibitory),0.9627659574468085,0.7075229850086789, +1e-07,glutamate sign: wholens (B excitatory),0.9183102371956979,0.7061936698801082, +1e-07,glutamate sign: rate_engine (B inhibitory),0.9916683534567922,0.6881621828584018, +1e-06,glutamate sign: wholens (B excitatory),0.9214816983710898,0.7503365230416149, +1e-06,glutamate sign: rate_engine (B inhibitory),0.9912807490852876,0.7251420547026872, +1e-05,glutamate sign: wholens (B excitatory),0.9541778975741239,0.7503567990386448, +1e-05,glutamate sign: rate_engine (B inhibitory),0.9919137466307278,0.7217133674537827, diff --git a/papers/tables/T25_engine_normalisation.md b/papers/tables/T25_engine_normalisation.md index ca80618..e8f5048 100644 --- a/papers/tables/T25_engine_normalisation.md +++ b/papers/tables/T25_engine_normalisation.md @@ -2,17 +2,17 @@ | conc_M | normalisation | spearman_rho_b_vs_d | reproduces_full_ordering | |---|---|---|---| -| 1.00e-06 | row_abs | 0.988 | True | -| 1.00e-06 | none | 0.906 | True | -| 1.00e-06 | degree | 0.666 | False | -| 1.00e-08 | row_abs | 0.906 | True | -| 1.00e-08 | none | 0.921 | True | -| 1.00e-08 | degree | -0.372 | False | -| 1.00e-08 | glutamate sign: wholens (B excitatory) | 0.906 | -- | -| 1.00e-08 | glutamate sign: rate_engine (B inhibitory) | 0.964 | -- | -| 1.00e-07 | glutamate sign: wholens (B excitatory) | 0.989 | -- | -| 1.00e-07 | glutamate sign: rate_engine (B inhibitory) | 0.987 | -- | -| 1.00e-06 | glutamate sign: wholens (B excitatory) | 0.988 | -- | -| 1.00e-06 | glutamate sign: rate_engine (B inhibitory) | 0.984 | -- | -| 1.00e-05 | glutamate sign: wholens (B excitatory) | 0.987 | -- | -| 1.00e-05 | glutamate sign: rate_engine (B inhibitory) | 0.979 | -- | +| 1.00e-06 | row_abs | 0.921 | True | +| 1.00e-06 | none | 0.935 | True | +| 1.00e-06 | degree | 0.551 | False | +| 1.00e-08 | row_abs | 0.876 | False | +| 1.00e-08 | none | 0.909 | True | +| 1.00e-08 | degree | -0.583 | False | +| 1.00e-08 | glutamate sign: wholens (B excitatory) | 0.876 | -- | +| 1.00e-08 | glutamate sign: rate_engine (B inhibitory) | 0.963 | -- | +| 1.00e-07 | glutamate sign: wholens (B excitatory) | 0.918 | -- | +| 1.00e-07 | glutamate sign: rate_engine (B inhibitory) | 0.992 | -- | +| 1.00e-06 | glutamate sign: wholens (B excitatory) | 0.921 | -- | +| 1.00e-06 | glutamate sign: rate_engine (B inhibitory) | 0.991 | -- | +| 1.00e-05 | glutamate sign: wholens (B excitatory) | 0.954 | -- | +| 1.00e-05 | glutamate sign: rate_engine (B inhibitory) | 0.992 | -- | diff --git a/papers/tables/T3_predictions.csv b/papers/tables/T3_predictions.csv index 53f49eb..b7c8554 100644 --- a/papers/tables/T3_predictions.csv +++ b/papers/tables/T3_predictions.csv @@ -3,6 +3,6 @@ H1,"At 1 uM, imidacloprid occupancy on insect nAChR is >= 0.9 while vertebrate a H2,The same dose lowers model drive through ACh edges in the MN9/DNp01 neighborhood: g_ach falls and the network mean rate falls with it.,subgraph,imidacloprid,1e-06,mean_hz,decrease,decrease,-6.145234834515873,-6.374205175471693,-5.916264493560052,-21.474928117137384,per,23,protocols/per.md,software_prediction, H3,Fipronil at 1 uM lowers g_gaba (RDL block) and disinhibits the neighborhood: the network mean rate rises above vehicle.,subgraph,fipronil,1e-06,mean_hz,increase,increase,0.8955929363577542,0.8622233051691006,0.9289625675464077,21.474928117137186,climbing,16,protocols/climbing.md,software_prediction, H4,"Diazepam occupies vertebrate GABA-A and does not occupy insect RDL in this library, so it moves no fly readout: the predicted circuit effect is exactly zero.",subgraph,diazepam,1e-06,mean_hz,none,none,0.0,0.0,0.0,,per,,protocols/per.md,software_prediction, -H5,Bitter input vetoes sugar-driven MN9 on the map-extracted labellar-GRN path (Shiu 2024 direction): adding bitter drive lowers MN9 below the sugar-only rate in vehicle.,taste_map,,0.0,mn9_sugar_bitter_hz - mn9_sugar_hz,decrease,decrease,-0.2958607006882387,-0.3041129980531392,-0.28760840332333826,-28.68698315218172,per,23,protocols/per.md,software_prediction, -H6,Fipronil at 1 uM relieves the bitter veto on the map path: the bitter veto ratio (MN9 sugar+bitter / MN9 sugar) rises above its vehicle value because RDL block removes the inhibition that carries the veto.,taste_map,fipronil,1e-06,bitter_veto_ratio,increase,increase,0.41291238946428677,0.40879827615134545,0.4170265027772281,80.30713471601884,per,16,protocols/per.md,software_prediction, -H7,"Picrotoxin is the non-selective control: it perturbs the fly neighborhood (RDL block raises the mean rate) while its receptor selectivity index is ~0, so any insect-selectivity claim must fail for it.",subgraph,picrotoxin,1e-06,mean_hz,increase,increase,0.3397826600493299,0.24404996599037576,0.435515354108284,2.8399676767899487,climbing,16,protocols/climbing.md,software_prediction, +H5,Bitter input vetoes sugar-driven MN9 on the map-extracted labellar-GRN path (Shiu 2024 direction): adding bitter drive lowers MN9 below the sugar-only rate in vehicle.,taste_map,,0.0,mn9_sugar_bitter_hz - mn9_sugar_hz,decrease,decrease,-0.2958607006882387,-0.3041129980531391,-0.2876084033233383,-28.686983152181778,per,23,protocols/per.md,software_prediction, +H6,Fipronil at 1 uM relieves the bitter veto on the map path: the bitter veto ratio (MN9 sugar+bitter / MN9 sugar) rises above its vehicle value because RDL block removes the inhibition that carries the veto.,taste_map,fipronil,1e-06,bitter_veto_ratio,increase,increase,0.41291238946428677,0.40879827615134545,0.4170265027772281,80.3071347160189,per,16,protocols/per.md,software_prediction, +H7,"Picrotoxin is the non-selective control: it perturbs the fly neighborhood (RDL block raises the mean rate) while its receptor selectivity index is ~0, so any insect-selectivity claim must fail for it.",subgraph,picrotoxin,1e-06,mean_hz,increase,increase,0.40019619368267084,0.2911695817476452,0.5092228056176965,2.9370605723918146,climbing,16,protocols/climbing.md,software_prediction, diff --git a/papers/tables/T3_predictions.md b/papers/tables/T3_predictions.md index 225c95b..594f9a6 100644 --- a/papers/tables/T3_predictions.md +++ b/papers/tables/T3_predictions.md @@ -8,4 +8,4 @@ | H4 | subgraph | diazepam | mean_hz | none | 0.000 | -- | software_prediction | | H5 | taste_map | -- | mn9_sugar_bitter_hz - mn9_sugar_hz | decrease | -0.296 | 23 | software_prediction | | H6 | taste_map | fipronil | bitter_veto_ratio | increase | 0.413 | 16 | software_prediction | -| H7 | subgraph | picrotoxin | mean_hz | increase | 0.340 | 16 | software_prediction | +| H7 | subgraph | picrotoxin | mean_hz | increase | 0.400 | 16 | software_prediction | diff --git a/papers/tables/T5_mixtures.csv b/papers/tables/T5_mixtures.csv index 9d8d447..0f1ec11 100644 --- a/papers/tables/T5_mixtures.csv +++ b/papers/tables/T5_mixtures.csv @@ -1,4 +1,4 @@ partner,components,published,published_as_flylab_verdict,model_verdict,null_model,matches,shared_receptors,observed_effect_fraction,expected_effect_fraction,combination_index,substituted -acephate (organophosphate),imidacloprid@1e-09; chlorpyrifos_oxon@1e-09,additive,additive,additive,bliss,True,,0.5971679900589311,0.5797876088693997,,"{""requested"": ""acephate"", ""used"": ""chlorpyrifos_oxon"", ""reason"": ""acephate is an organophosphate pro-insecticide acting through AChE inhibition; the library's only AChE inhibitor is chlorpyrifos-oxon""}" +acephate (organophosphate),imidacloprid@1e-09; chlorpyrifos_oxon@1e-09,additive,additive,additive,bliss,True,,0.010947376008838315,0.010947376008838328,,"{""requested"": ""acephate"", ""used"": ""chlorpyrifos_oxon"", ""reason"": ""acephate is an organophosphate pro-insecticide acting through AChE inhibition; the library's only AChE inhibitor is chlorpyrifos-oxon""}" lambda-cyhalothrin (pyrethroid),imidacloprid@1e-09; deltamethrin@1e-09,additive,additive,additive,bliss,True,,0.08202536675147733,0.08035211193101388,,"{""requested"": ""lambda-cyhalothrin"", ""used"": ""deltamethrin"", ""reason"": ""same class (type II pyrethroid, insect_Nav); lambda-cyhalothrin is not in the FlyLab library""}" -clothianidin (neonicotinoid),imidacloprid@1e-09; clothianidin@1e-09,no_interaction,additive,additive,loewe,True,insect_nAChR; vertebrate_nAChR_a4b2,0.016187836664300598,0.017970764065576184,1.0589850973099812, +clothianidin (neonicotinoid),imidacloprid@1e-09; clothianidin@1e-09,no_interaction,additive,additive,loewe,True,insect_nAChR; vertebrate_nAChR_a4b2,0.01618783666430054,0.017970764065576184,1.0589850973099812, diff --git a/papers/tables/T6_dependence_profile.md b/papers/tables/T6_dependence_profile.md index 17cdc61..7a59746 100644 --- a/papers/tables/T6_dependence_profile.md +++ b/papers/tables/T6_dependence_profile.md @@ -1,4 +1,4 @@ -**T6_dependence_profile.** Connectome-dependence profile. `real_effect` is treated minus vehicle on the real MaleCNS cut; the null is the same contrast on `n` degraded copies of that cut, with the seed block and node order held fixed. `p_two_sided` is the empirical permutation probability (k+1)/(n+1) and is the statistic to read first; `p_resolution` is its floor, and `z` is a standardised distance from a usually non-normal null. `verdict` is the only claim the design supports: `distinguishable` when the test rejects, `equivalent_within_tolerance` when the gap from the null median is below the prespecified margin `delta`, and `indeterminate` otherwise -- a failure to reject is never evidence of equivalence. `sign_permute` is a **joint** target-set-and-sign null and is off the ladder (`on_ladder = False`); the rank-3 rung is the weight-matched transmitter null, which holds each transmitter's share of total outgoing weight fixed (T23). +**T6_dependence_profile.** Connectome-dependence profile. `real_effect` is treated minus vehicle on the real MaleCNS cut; the null is the same contrast on `n` degraded copies of that cut, with the seed block and node order held fixed. `p_two_sided` is the empirical permutation probability (k+1)/(n+1) and is the statistic to read first; `p_resolution` is its floor, and `z` is a standardised distance from a usually non-normal null. `verdict` is the only claim the design supports: `distinguishable` when the test rejects, `equivalent_within_tolerance` when the gap from the null median is below the prespecified margin `delta`, and `indeterminate` otherwise -- the point-gap rule is not a formal equivalence test. `sign_permute` is a **joint** target-set-and-sign null and is off the ladder (`on_ladder = False`); the rank-3 rung is the weight-matched transmitter null, which holds each transmitter's share of total outgoing weight fixed (T23). | compound | assay | mode | on_ladder | real_effect | p_two_sided | verdict | abs_gap_from_null_median | delta | |---|---|---|---|---|---|---|---|---| diff --git a/papers/tables/T8_landscape.csv b/papers/tables/T8_landscape.csv index dcf47a2..441f1a4 100644 --- a/papers/tables/T8_landscape.csv +++ b/papers/tables/T8_landscape.csv @@ -7,14 +7,14 @@ caffeine,methylxanthine,named,,,,,,,,0.6188311115034872,0.3007857332038492,every caffeine,methylxanthine,taste_motor,,,,,,,,0.5868443687762142,0.30350172470360814,every insect receptor row for this compound is a class placeholder: it cannot move the fly circuit by construction chlordimeform,formamidine,named,,,,,,,0.0,0.6188311115034872,0.3007857332038492, chlordimeform,formamidine,taste_motor,,,,,,,0.0,0.5868443687762142,0.30350172470360814, -chlorpyrifos_oxon,organophosphate (oxon metabolite),named,AChE,1.4771212547196624,8.096517267532963e-10,7.5e-08,1.9667630169519281,0.48964176223226574,6.3170756098203755,0.6188311115034872,0.3007857332038492, -chlorpyrifos_oxon,organophosphate (oxon metabolite),taste_motor,AChE,1.4771212547196624,6.201965662788153e-10,7.5e-08,2.082531905957163,0.6054106512375006,7.8783253122676165,0.5868443687762142,0.30350172470360814, +chlorpyrifos_oxon,organophosphate (oxon metabolite),named,,,,,,,,0.6188311115034872,0.3007857332038492,every insect receptor row for this compound is a class placeholder: it cannot move the fly circuit by construction +chlorpyrifos_oxon,organophosphate (oxon metabolite),taste_motor,,,,,,,,0.5868443687762142,0.30350172470360814,every insect receptor row for this compound is a class placeholder: it cannot move the fly circuit by construction clothianidin,neonicotinoid,named,nAChR,3.522878745280338,5.921818922734351e-08,2.5e-05,2.625484885264079,-0.8973938600162588,0.9308633342364611,0.6188311115034872,0.3007857332038492, clothianidin,neonicotinoid,taste_motor,nAChR,3.522878745280338,6.03772126641474e-08,2.5e-05,2.61706694888396,-0.9058117963963777,0.8525710945472482,0.5868443687762142,0.30350172470360814, -ddt,organochlorine,named,Nav,1.1870866433571445,1.0722707580748198e-05,0.00025,1.3676355461786587,0.18054890282151415,2.9243296089441526,0.6188311115034872,0.3007857332038492, +ddt,organochlorine,named,Nav,1.1870866433571445,1.0722707580748198e-05,0.00025,1.3676355461786587,0.18054890282151415,2.924329608944152,0.6188311115034872,0.3007857332038492, ddt,organochlorine,taste_motor,Nav,1.1870866433571445,1.1043426929115513e-05,0.00025,1.3548361467406957,0.1677495033835512,1.945115994189779,0.5868443687762142,0.30350172470360814, -deltamethrin,type II pyrethroid,named,Nav,2.0655015487564325,6.5366894677292685e-09,1.25e-06,2.281552159165738,0.2160506104093054,5.755973991100234,0.6188311115034872,0.3007857332038492, -deltamethrin,type II pyrethroid,taste_motor,Nav,2.0655015487564325,6.7624488955838246e-09,1.25e-06,2.2668060168823114,0.20130446812587888,3.8248137750907354,0.5868443687762142,0.30350172470360814, +deltamethrin,type II pyrethroid,named,Nav,2.065501548756432,6.5366894677292685e-09,1.25e-06,2.281552159165738,0.21605061040930584,5.755973991100234,0.6188311115034872,0.3007857332038492, +deltamethrin,type II pyrethroid,taste_motor,Nav,2.065501548756432,6.7624488955838246e-09,1.25e-06,2.2668060168823114,0.20130446812587932,3.8248137750907354,0.5868443687762142,0.30350172470360814, diazepam,benzodiazepine,named,,,,,,,,0.6188311115034872,0.3007857332038492,every insect receptor row for this compound is a class placeholder: it cannot move the fly circuit by construction diazepam,benzodiazepine,taste_motor,,,,,,,,0.5868443687762142,0.30350172470360814,every insect receptor row for this compound is a class placeholder: it cannot move the fly circuit by construction dieldrin,cyclodiene organochlorine,named,GABA_A,1.0,,2.5e-07,,,0.13566196399429847,0.6188311115034872,0.3007857332038492, @@ -25,19 +25,19 @@ gaba,endogenous transmitter,named,GABA_A,-0.6989700043360187,,3.968502629920499e gaba,endogenous transmitter,taste_motor,GABA_A,-0.6989700043360187,,3.968502629920499e-06,,,0.0694065018292579,0.5868443687762142,0.30350172470360814, imidacloprid,neonicotinoid,named,nAChR,2.6989700043360187,4.0210203852332524e-08,2.5e-06,1.793603733835261,-0.9053662705007577,0.9308633342364611,0.6188311115034872,0.3007857332038492, imidacloprid,neonicotinoid,taste_motor,nAChR,2.6989700043360187,4.0526909833436914e-08,2.5e-06,1.7901965185116522,-0.9087734858243666,0.8525710945472482,0.5868443687762142,0.30350172470360814, -ivermectin,avermectin,named,GluCl,1.6989700043360187,,2.5e-07,,,0.04809050763205058,0.6188311115034872,0.3007857332038492, -ivermectin,avermectin,taste_motor,GluCl,1.6989700043360187,,2.5e-07,,,0.09246981630646477,0.5868443687762142,0.30350172470360814, +ivermectin,avermectin,named,GluCl,1.6989700043360187,,2.5e-07,,,0.0056401363945177444,0.6188311115034872,0.3007857332038492, +ivermectin,avermectin,taste_motor,GluCl,1.6989700043360187,,2.5e-07,,,0.0037118955059856095,0.5868443687762142,0.30350172470360814, nicotine,alkaloid,named,nAChR,0.6989700043360189,4.2666353973946507e-07,3.149802624737183e-07,-0.1318021915270684,-0.8307721958630873,0.9308633342364611,0.6188311115034872,0.3007857332038492, nicotine,alkaloid,taste_motor,nAChR,0.6989700043360189,4.304638098448084e-07,3.149802624737183e-07,-0.13565330455514335,-0.8346233088911622,0.8525710945472482,0.5868443687762142,0.30350172470360814, nitenpyram,neonicotinoid,named,nAChR,2.6020599913279625,9.753588971361666e-08,5e-06,1.70980555441641,-0.8922544369115526,0.9308633342364611,0.6188311115034872,0.3007857332038492, nitenpyram,neonicotinoid,taste_motor,nAChR,2.6020599913279625,9.630935367370797e-08,5e-06,1.7153015359902444,-0.8867584553377181,0.8525710945472482,0.5868443687762142,0.30350172470360814, -permethrin,type I pyrethroid,named,Nav,2.0,6.004762270454635e-08,1e-05,2.221504181712107,0.22150418171210706,5.754773828305756,0.6188311115034872,0.3007857332038492, -permethrin,type I pyrethroid,taste_motor,Nav,2.0,6.169253473111401e-08,1e-05,2.2097673857436817,0.2097673857436817,3.8203989947836776,0.5868443687762142,0.30350172470360814, +permethrin,type I pyrethroid,named,,,6.004762270454635e-08,,,,5.754773828305756,0.6188311115034872,0.3007857332038492, +permethrin,type I pyrethroid,taste_motor,,,6.169253473111401e-08,,,,3.8203989947836776,0.5868443687762142,0.30350172470360814, picrotoxin,plant convulsant,named,GABA_A,0.0,,2.5e-07,,,0.13566196399429847,0.6188311115034872,0.3007857332038492, picrotoxin,plant convulsant,taste_motor,GABA_A,0.0,,2.5e-07,,,0.4143743777977133,0.5868443687762142,0.30350172470360814, -spinosad,spinosyn,named,,,1.1352453295350782e-05,,,,0.9308633342364611,0.6188311115034872,0.3007857332038492, -spinosad,spinosyn,taste_motor,,,1.126405886074481e-05,,,,0.8525710945472482,0.5868443687762142,0.30350172470360814, -sulfoxaflor,sulfoximine,named,nAChR,0.8239087409443188,,0.00025,,,0.10682866262492406,0.6188311115034872,0.3007857332038492, -sulfoxaflor,sulfoximine,taste_motor,nAChR,0.8239087409443188,,0.00025,,,0.11510842458122944,0.5868443687762142,0.30350172470360814, -thiamethoxam,neonicotinoid,named,,,,,,,0.21682214846276826,0.6188311115034872,0.3007857332038492, -thiamethoxam,neonicotinoid,taste_motor,,,,,,,0.22998613679206625,0.5868443687762142,0.30350172470360814, +spinosad,spinosyn,named,,,,,,,,0.6188311115034872,0.3007857332038492,every insect receptor row for this compound is a class placeholder: it cannot move the fly circuit by construction +spinosad,spinosyn,taste_motor,,,,,,,,0.5868443687762142,0.30350172470360814,every insect receptor row for this compound is a class placeholder: it cannot move the fly circuit by construction +sulfoxaflor,sulfoximine,named,,,,,,,,0.6188311115034872,0.3007857332038492,every insect receptor row for this compound is a class placeholder: it cannot move the fly circuit by construction +sulfoxaflor,sulfoximine,taste_motor,,,,,,,,0.5868443687762142,0.30350172470360814,every insect receptor row for this compound is a class placeholder: it cannot move the fly circuit by construction +thiamethoxam,neonicotinoid,named,,,,,,,0.2168221484627684,0.6188311115034872,0.3007857332038492, +thiamethoxam,neonicotinoid,taste_motor,,,,,,,0.2299861367920664,0.5868443687762142,0.30350172470360814, diff --git a/papers/tables/T8_landscape.md b/papers/tables/T8_landscape.md index 626da53..f8fa35c 100644 --- a/papers/tables/T8_landscape.md +++ b/papers/tables/T8_landscape.md @@ -10,8 +10,8 @@ | caffeine | methylxanthine | taste_motor | -- | -- | -- | | chlordimeform | formamidine | named | -- | -- | -- | | chlordimeform | formamidine | taste_motor | -- | -- | -- | -| chlorpyrifos_oxon | organophosphate (oxon metabolite) | named | 1.477 | 1.967 | 0.490 | -| chlorpyrifos_oxon | organophosphate (oxon metabolite) | taste_motor | 1.477 | 2.083 | 0.605 | +| chlorpyrifos_oxon | organophosphate (oxon metabolite) | named | -- | -- | -- | +| chlorpyrifos_oxon | organophosphate (oxon metabolite) | taste_motor | -- | -- | -- | | clothianidin | neonicotinoid | named | 3.523 | 2.625 | -0.897 | | clothianidin | neonicotinoid | taste_motor | 3.523 | 2.617 | -0.906 | | ddt | organochlorine | named | 1.187 | 1.368 | 0.181 | @@ -34,13 +34,13 @@ | nicotine | alkaloid | taste_motor | 0.699 | -0.136 | -0.835 | | nitenpyram | neonicotinoid | named | 2.602 | 1.710 | -0.892 | | nitenpyram | neonicotinoid | taste_motor | 2.602 | 1.715 | -0.887 | -| permethrin | type I pyrethroid | named | 2.000 | 2.222 | 0.222 | -| permethrin | type I pyrethroid | taste_motor | 2.000 | 2.210 | 0.210 | +| permethrin | type I pyrethroid | named | -- | -- | -- | +| permethrin | type I pyrethroid | taste_motor | -- | -- | -- | | picrotoxin | plant convulsant | named | 0.000 | -- | -- | | picrotoxin | plant convulsant | taste_motor | 0.000 | -- | -- | | spinosad | spinosyn | named | -- | -- | -- | | spinosad | spinosyn | taste_motor | -- | -- | -- | -| sulfoxaflor | sulfoximine | named | 0.824 | -- | -- | -| sulfoxaflor | sulfoximine | taste_motor | 0.824 | -- | -- | +| sulfoxaflor | sulfoximine | named | -- | -- | -- | +| sulfoxaflor | sulfoximine | taste_motor | -- | -- | -- | | thiamethoxam | neonicotinoid | named | -- | -- | -- | | thiamethoxam | neonicotinoid | taste_motor | -- | -- | -- | diff --git a/pyproject.toml b/pyproject.toml index 5a69029..e6cf1b2 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -23,6 +23,7 @@ dependencies = [ [project.optional-dependencies] dev = ["pytest>=8.0", "httpx>=0.27"] viz = ["matplotlib>=3.8"] +browser-test = ["playwright>=1.48"] # The science core only needs these two. This is what the static (Pyodide) # build installs; fastapi/pydantic/typer/pandas/pyarrow are server and CLI # dependencies and are never loaded in the browser -- see docs/PAGES.md. diff --git a/scripts/reanalyze_vnc_expression.py b/scripts/reanalyze_vnc_expression.py new file mode 100644 index 0000000..a05786e --- /dev/null +++ b/scripts/reanalyze_vnc_expression.py @@ -0,0 +1,239 @@ +#!/usr/bin/env python3 +"""Recompute VNC-wide target-gene detection from Allen et al. GEO matrices. + +This is a cell-level transcript-detection summary, not a cell-type annotation. +It deliberately does not feed these pooled VNC rates into FlyLab's +MaleCNS-superclass gains. The GEO archive contains four 30,000-barcode DGE +matrices; the published nGene, nUMI and mitochondrial cutoffs are applied +before any detection fraction is calculated. + + python scripts/reanalyze_vnc_expression.py /path/to/GSE141807_RAW.tar +""" +from __future__ import annotations + +import argparse +import gzip +import hashlib +import re +import tarfile +from pathlib import Path +from typing import Any, BinaryIO + +import numpy as np +import yaml + +ARCHIVE_SHA256 = "bda31982a53a432bb4712ebdaf11ac7d8b9e7be580f72c2acaf7a58912e64daf" +GEO_ACCESSION = "GSE141807" +EXPECTED_MEMBERS = { + "GSM4213596_female_rep1_dge.txt.gz": ("GSM4213596", "female", 2590), + "GSM4213597_female_rep2_dge.txt.gz": ("GSM4213597", "female", 9060), + "GSM4213598_male_rep1_dge.txt.gz": ("GSM4213598", "male", 6522), + "GSM4213599_male_rep2_dge.txt.gz": ("GSM4213599", "male", 8596), +} + +# These are gene-name rules aligned to the expression layer's coarse target +# keys. The Ace group is an acetylcholinesterase target, not a receptor. +TARGET_PREFIXES: dict[str, tuple[re.Pattern[str], ...]] = { + "insect_nAChR": (re.compile(r"^nAChR", re.I),), + "insect_RDL": (re.compile(r"^Rdl$", re.I),), + "insect_GluCl": (re.compile(r"^GluCl", re.I),), + "insect_OctR": (re.compile(r"^(OctR|Oct)", re.I),), + "insect_AChE": (re.compile(r"^Ace", re.I),), +} + +PAPER_QC = { + "nGene_gt": 200, + "nUMI_gt": 1200, + "nUMI_lte": 10000, + "prop_mito_lte": 0.15, +} + + +def _sha256(path: Path) -> str: + digest = hashlib.sha256() + with path.open("rb") as handle: + for block in iter(lambda: handle.read(1024 * 1024), b""): + digest.update(block) + return digest.hexdigest() + + +def analyze_matrix(stream: BinaryIO, member_name: str) -> dict[str, Any]: + """Analyze one tab-separated genes-by-barcodes matrix from GEO's tar.""" + try: + sample_id, sex, published_cells = EXPECTED_MEMBERS[member_name] + except KeyError as exc: + raise ValueError(f"unexpected GSE141807 matrix member: {member_name}") from exc + + header = stream.readline().rstrip(b"\r\n").split(b"\t") + if not header or header[0] != b"GENE": + raise ValueError(f"{member_name}: expected GENE as the first matrix header") + n_cells = len(header) - 1 + if n_cells <= 0: + raise ValueError(f"{member_name}: matrix contains no barcodes") + + n_umi = np.zeros(n_cells, dtype=np.int64) + n_gene = np.zeros(n_cells, dtype=np.int32) + n_mito = np.zeros(n_cells, dtype=np.int64) + family_detected = {key: np.zeros(n_cells, dtype=bool) for key in TARGET_PREFIXES} + gene_detected: dict[str, np.ndarray] = {} + mitochondrial_genes: set[str] = set() + matrix_genes: set[str] = set() + n_rows = 0 + + for line_number, line in enumerate(stream, start=2): + line = line.rstrip(b"\r\n") + gene_bytes, separator, counts_bytes = line.partition(b"\t") + if not separator: + raise ValueError(f"{member_name}:{line_number}: row has no count columns") + gene = gene_bytes.decode("utf-8", "replace") + if gene in matrix_genes: + raise ValueError(f"{member_name}:{line_number}: duplicate gene symbol {gene!r}") + matrix_genes.add(gene) + counts = np.fromstring(counts_bytes, sep="\t", dtype=np.int32) + if counts.size != n_cells: + raise ValueError( + f"{member_name}:{line_number}: expected {n_cells} counts, found {counts.size}" + ) + if np.any(counts < 0): + raise ValueError(f"{member_name}:{line_number}: negative UMI count") + + n_rows += 1 + n_umi += counts + detected = counts > 0 + n_gene += detected + if gene.startswith("mt:"): + n_mito += counts + mitochondrial_genes.add(gene) + + for key, patterns in TARGET_PREFIXES.items(): + if any(pattern.match(gene) for pattern in patterns): + family_detected[key] |= detected + if gene not in gene_detected: + gene_detected[gene] = detected.copy() + else: # keep duplicate gene rows conservative if source changes + gene_detected[gene] |= detected + + prop_mito = n_mito / np.maximum(n_umi, 1) + keep = ( + (n_gene > PAPER_QC["nGene_gt"]) + & (n_umi > PAPER_QC["nUMI_gt"]) + & (n_umi <= PAPER_QC["nUMI_lte"]) + & (prop_mito <= PAPER_QC["prop_mito_lte"]) + ) + kept = int(keep.sum()) + target_summary: dict[str, Any] = {} + for key in TARGET_PREFIXES: + genes = { + gene: { + "n_detected": int(mask[keep].sum()), + "fraction_detected": float(mask[keep].mean()) if kept else None, + } + for gene, mask in sorted(gene_detected.items()) + if any(pattern.match(gene) for pattern in TARGET_PREFIXES[key]) + } + target_summary[key] = { + "n_genes_matched": len(genes), + "genes": genes, + "n_cells_with_any_matched_gene_detected": int(family_detected[key][keep].sum()), + "fraction_cells_with_any_matched_gene_detected": ( + float(family_detected[key][keep].mean()) if kept else None + ), + } + + return { + "sample_id": sample_id, + "sex": sex, + "cells_in_matrix": n_cells, + "genes_in_matrix": n_rows, + "cells_after_published_qc": kept, + "published_cells_after_qc": published_cells, + "cell_count_difference": kept - published_cells, + "mitochondrial_genes": sorted(mitochondrial_genes), + "targets": target_summary, + } + + +def analyze_archive(archive: Path) -> dict[str, Any]: + """Validate the source archive and emit its four per-replicate summaries.""" + checksum = _sha256(archive) + if checksum != ARCHIVE_SHA256: + raise ValueError( + f"unexpected GEO archive SHA-256 {checksum}; expected {ARCHIVE_SHA256}" + ) + + results: dict[str, Any] = {} + with tarfile.open(archive, mode="r:") as tar: + members = {Path(member.name).name: member for member in tar.getmembers() if member.isfile()} + if set(members) != set(EXPECTED_MEMBERS): + raise ValueError( + "unexpected archive contents: " + ", ".join(sorted(members)) + ) + for name in EXPECTED_MEMBERS: + member = members[name] + raw = tar.extractfile(member) + if raw is None: + raise ValueError(f"cannot read {name} from GEO archive") + with gzip.GzipFile(fileobj=raw, mode="rb") as matrix: + results[EXPECTED_MEMBERS[name][0]] = analyze_matrix(matrix, name) + + all_cells = sum(row["cells_after_published_qc"] for row in results.values()) + all_published = sum(row["published_cells_after_qc"] for row in results.values()) + return { + "schema_version": 1, + "source": { + "accession": GEO_ACCESSION, + "archive": "GSE141807_RAW.tar", + "archive_sha256": checksum, + "geo_url": "https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE141807", + "paper_url": "https://elifesciences.org/articles/54074", + }, + "method": { + "scope": "adult VNC-wide transcript detection; cells are not assigned to cell classes", + "qc_thresholds_from_paper": PAPER_QC, + "detection_rule": "one or more UMI for the gene in a QC-passing barcode", + "dropout_note": "a zero count is not evidence that a cell does not express the gene", + "not_used_for": "MaleCNS superclass gains, receptor-complex inference, or MN9 expression", + }, + "cells_after_qc_total": all_cells, + "published_cells_after_qc_total": all_published, + "all_replicate_cell_counts_match_paper": all( + row["cell_count_difference"] == 0 for row in results.values() + ), + "cell_count_reproduction_note": ( + "Three replicate counts match the paper. GSM4213598 retains 6,524 cells versus " + "6,522 reported (difference +2) under the cited QC thresholds; this small " + "discrepancy is unresolved and was not adjusted away." + ), + "replicates": results, + "interpretation": ( + "Descriptive VNC-wide RNA detection only. Detection fractions are not receptor " + "complex prevalence or cell-type-specific expression fractions." + ), + } + + +def main(argv: list[str] | None = None) -> int: + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument("archive", type=Path, help="GSE141807_RAW.tar downloaded from NCBI GEO") + parser.add_argument( + "--out", + type=Path, + default=Path("data/literature/vnc_receptor_detection_GSE141807.yaml"), + help="output summary YAML (default: %(default)s)", + ) + args = parser.parse_args(argv) + result = analyze_archive(args.archive) + args.out.parent.mkdir(parents=True, exist_ok=True) + args.out.write_text(yaml.safe_dump(result, sort_keys=False, allow_unicode=True)) + print(f"wrote {args.out}") + for row in result["replicates"].values(): + print( + f" {row['sample_id']} ({row['sex']}): " + f"{row['cells_after_published_qc']:,} retained; " + f"published {row['published_cells_after_qc']:,}" + ) + return 0 + + +if __name__ == "__main__": + raise SystemExit(main()) diff --git a/scripts/reproduce_paper.py b/scripts/reproduce_paper.py index 93857f4..67f28b7 100644 --- a/scripts/reproduce_paper.py +++ b/scripts/reproduce_paper.py @@ -361,13 +361,20 @@ def save_table( fields: Iterable[str], caption: str, md_fields: Iterable[str] | None = None, + *, + lineterminator: str = "\r\n", ) -> None: """Write ``name.csv`` (all fields) and ``name.md`` (a readable subset).""" ctx.ensure_dirs() fields = list(fields) csv_path = ctx.tabdir / f"{name}.csv" with csv_path.open("w", newline="") as fh: - w = csv.DictWriter(fh, fieldnames=fields, extrasaction="ignore") + w = csv.DictWriter( + fh, + fieldnames=fields, + extrasaction="ignore", + lineterminator=lineterminator, + ) w.writeheader() for r in rows: w.writerow({k: _csv_cell(r.get(k)) for k in fields}) @@ -755,15 +762,16 @@ def step_scorecard(ctx: Ctx) -> None: fig, "F2_dual_scorecard", f"Dual scorecard for imidacloprid and fipronil at {_fmt_M(PAPER_CONC)}. Bars are " - "Hill occupancy of the insect target (blue) and its vertebrate counterpart " + "Hill-form engagement of the insect target (blue) and its vertebrate counterpart " "(orange) for each of the five receptor pairs declared in " "`library.yaml:selectivity_pairs`. Hatched bars rest on a class placeholder: " "the compound has no sourced EC50 at that pair and the row is inert. " f"Imidacloprid reaches {ctx.get('imi_insect_occ'):.3f} at insect nAChR against " f"{ctx.get('imi_vert_occ'):.3f} at vertebrate a4b2; fipronil reaches " f"{ctx.get('fip_insect_occ'):.3f} at insect RDL against {ctx.get('fip_vert_occ'):.3f} " - "at native heteromeric vertebrate GABA-A. Teaching-tier EC50s: these are " - "literature-order values, not measured constants.", + "at a multi-species vertebrate GABA-related binding-site proxy. These are " + "heterogeneous assay inputs and transfers, not measured fractional occupancy " + "of matched receptors in living animals.", ) @@ -830,14 +838,14 @@ def step_curves(ctx: Ctx) -> None: for ax in axes[-1]: ax.set_xlabel("free concentration (M)") for ax in axes[:, 0]: - ax.set_ylabel("fractional occupancy") - fig.suptitle("Occupancy curves and insect-over-vertebrate selectivity windows", y=0.98) + ax.set_ylabel("Hill-form engagement proxy") + fig.suptitle("Model engagement curves and insect-over-vertebrate selectivity windows", y=0.98) fig.tight_layout() save_fig( ctx, fig, "F3_occupancy_curves", - "Hill occupancy against free concentration for the four-compound panel, one " + "Hill-form model engagement against free concentration for the four-compound panel, one " "insect target (solid, blue) and its vertebrate counterpart (dashed, orange) " "per compound. The shaded band is the selectivity window between the two " "EC50s, annotated in log10 units; the dotted vertical line is the " @@ -1014,7 +1022,7 @@ def step_graph(ctx: Ctx) -> None: "node that itself receives input (mean degree %.2f). A degree-preserving " "rewire of such a graph is close to the identity, so a negative topology " "verdict on this cut is weak evidence. The `taste_motor` cut (mean degree " - "%.2f) is the substrate to generalise from." + "%.2f) is the denser comparison substrate." % ( ctx.text("named_share_onto_seeds"), ctx.get("named_edges") or 0, @@ -1264,8 +1272,9 @@ def step_veto(ctx: Ctx) -> None: r = nb["readouts"] data[engine][label] = r tag = f"{engine}_{label}" - ctx.put(f"veto_{tag}_sugar", _f(r["mn9_sugar_hz"]), unit="Hz") - ctx.put(f"veto_{tag}_sugar_bitter", _f(r["mn9_sugar_bitter_hz"]), unit="Hz") + output_unit = "rate-model units" if engine == "rate" else "Hz" + ctx.put(f"veto_{tag}_sugar", _f(r["mn9_sugar_hz"]), unit=output_unit) + ctx.put(f"veto_{tag}_sugar_bitter", _f(r["mn9_sugar_bitter_hz"]), unit=output_unit) ctx.put(f"veto_{tag}_ratio", _f(r["bitter_veto_ratio"])) ctx.put("veto_n_sweet_grn", int(data["rate"]["vehicle"]["n_sweet_grn"])) ctx.put("veto_n_bitter_grn", int(data["rate"]["vehicle"]["n_bitter_grn"])) @@ -1274,7 +1283,8 @@ def step_veto(ctx: Ctx) -> None: if data["rate"]["imidacloprid"]["bitter_veto_ratio"] is None: ctx.note( "imidacloprid at 1 uM silences MN9 on the map path on both engines " - "(sugar-driven MN9 = 0 Hz), so its bitter-veto ratio is undefined, not 0. " + "(sugar-driven MN9 = 0 in each engine's output units), so its bitter-veto " + "ratio is undefined, not 0. " "g_ach hits the 0.05 floor, which removes the cholinergic GRN->MN9 drive " "itself. This is a property of the agonist patch rule, and it is why the " "taste-map arm of the null panel is undefined for imidacloprid." @@ -1302,7 +1312,7 @@ def step_veto(ctx: Ctx) -> None: ) ax.set_xticks(x) ax.set_xticklabels(labels) - ax.set_ylabel("MN9 rate (Hz)") + ax.set_ylabel("MN9 output (rate-model units)" if engine == "rate" else "MN9 rate (Hz)") ax.set_title(f"{engine} engine") ax.set_ylim(0, max(sugar + both) * 1.3 + 0.1) ax.grid(axis="y", lw=0.5) @@ -1325,7 +1335,7 @@ def step_veto(ctx: Ctx) -> None: f"{_fmt_M(PAPER_CONC)} blocks RDL, raises the sugar-only rate and partially " f"relieves the veto (ratio {ctx.get('veto_rate_fipronil_ratio'):.2f} rate, " f"{ctx.get('veto_lif_fipronil_ratio'):.2f} LIF). Imidacloprid silences MN9 " - "altogether, so its ratio is undefined rather than zero. Absolute rates differ " + "altogether, so its ratio is undefined rather than zero. Outputs differ " "between engines by construction and are not calibrated to any recording.", ) @@ -1605,7 +1615,8 @@ def step_evidence(ctx: Ctx) -> None: fig, "F11_evidence_types", "What the compound library rests on, row by row. Left: the parameter each " - "cited source actually measured. Middle: the ordered evidence distance between " + "library row declares. Some source-to-value links remain under assay-level " + "review. Middle: the declared evidence distance between " "that source and this compound at this receptor in this species. Right: the " "transformation the type system then permits, which is a function of the two. " f"Green marks the {ctx.get('ev_dist_E0')} on-target (E0) rows and the " @@ -1615,8 +1626,8 @@ def step_evidence(ctx: Ctx) -> None: "gap; red marks rows that carry no modellable evidence and return N/A rather " "than a small number. The type system, not a convention, enforces this: asking " "for an engagement from an untyped row raises rather than returning a value, " - "and a soundness invariant is checked exhaustively over every row and every " - "entry point.", + "and the admission invariant is checked over shipped rows and enumerated " + "entry points. This does not verify the biological validity of admitted values.", ) @@ -1641,10 +1652,10 @@ def step_dependence(ctx: Ctx) -> None: is an in-star (T19); ``taste_motor`` has mean degree 10.4, and the two do not agree. - *The instrument is tested.* A known topology-dependent effect is planted - in a synthetic cut of the same size and density and the ladder is asked to - find it, with a power surface over planted strength and permutation count - (T22, F16). + *The instrument has limited positive controls.* A known topology-dependent + effect is planted in a 250-node synthetic cut that is not matched to the + real extract, and detection is counted over planted strengths and + permutation budgets (T22, F16). """ import numpy as np @@ -1657,6 +1668,7 @@ def step_dependence(ctx: Ctx) -> None: dependence_profile, ladder_power, ladder_recovery, + _local_null_draws, ) from flylab.analysis.nullmodels import MODES, null_distribution @@ -1745,13 +1757,13 @@ def arm(compound: str, assay: str, readout: str, n_i: int, tag: str) -> dict[str ctx.put( f"dep_{tag}_{compound}_effect", _f(prof["real_effect"]), - unit="Hz" if assay == "subgraph" else None, + unit="rate-model units" if assay == "subgraph" else None, ) ctx.put(f"dep_{tag}_{compound}_vehicle", _f(prof.get("real_vehicle"))) ctx.put(f"dep_{tag}_{compound}_class", cls["class"]) ctx.put(f"dep_{tag}_{compound}_level", lvl.get("level")) ctx.put(f"dep_{tag}_{compound}_level_verdict", lvl.get("verdict")) - ctx.put(f"dep_{tag}_{compound}_delta", _f(prof.get("delta")), unit="Hz") + ctx.put(f"dep_{tag}_{compound}_delta", _f(prof.get("delta")), unit="rate-model units") ctx.put( f"dep_{tag}_{compound}_delta_frac", _f(prof.get("delta_frac")), @@ -1781,7 +1793,8 @@ def arm(compound: str, assay: str, readout: str, n_i: int, tag: str) -> dict[str min([p for p in struct if p is not None], default=None), ) - # draws and prefix sweep, for F6 and T11 only + # Draws and prefix sweep for F6 and T11. The balanced transmitter + # null is local to dependence.py, so it needs its own draw path. for mode in MODES: d = null_distribution( assay, @@ -1810,6 +1823,16 @@ def arm(compound: str, assay: str, readout: str, n_i: int, tag: str) -> dict[str ctx.put(f"sweep_{tag}_{compound}_{mode}_n{cp['n']}", _f(cp["p"])) if assay == "subgraph": draws[(compound, mode)] = [v for v in d["null_effects"] if v is not None] + if assay == "subgraph": + matched_draws, _ = _local_null_draws( + assay=assay, compound=compound, conc_M=PAPER_CONC, + readout=readout, graph_name="named", + mode="sign_permute_weight_matched", n=int(n_i), + seed=int(ctx.seed), kw={}, + ) + draws[(compound, "sign_permute_weight_matched")] = [ + v for v in matched_draws if v is not None + ] return prof named = {c: arm(c, "subgraph", "mean_hz", n, "named") for c in ("imidacloprid", "fipronil")} @@ -1823,10 +1846,10 @@ def arm(compound: str, assay: str, readout: str, n_i: int, tag: str) -> dict[str ) if gap_plain is not None and gap_matched is not None: ctx.note( - "the transmitter null matters to the equivalence claim and not to the " - "probability: imidacloprid's gap from the null median is %.2f Hz under the " - "plain label permutation and %.2f Hz under the weight-matched null, against " - "a prespecified margin of %s Hz. The plain permutation is a joint " + "the transmitter null changes the descriptive point-gap verdict: " + "imidacloprid's gap from the null median is %.2f rate-model units under " + "the plain label permutation and %.2f under the weight-matched null, " + "against a declared margin of %s rate-model units. The plain permutation is a joint " "target-set-and-sign null and is reported as one." % (gap_plain, gap_matched, ctx.text("dep_named_imidacloprid_delta")) ) @@ -1835,7 +1858,7 @@ def arm(compound: str, assay: str, readout: str, n_i: int, tag: str) -> dict[str "the fipronil bitter-veto ratio on the taste-motor path is a negative at " "n=%d shuffles (max |z| = %s, p resolution %.4f): the veto ratio is model " "behaviour, not evidence about the wiring, and the verdicts say which " - "modes are equivalent within tolerance (%s) and which are merely " + "modes pass the descriptive point-gap rule (%s) and which are merely " "indeterminate (%s)." % ( n_taste, @@ -1882,8 +1905,8 @@ def arm(compound: str, assay: str, readout: str, n_i: int, tag: str) -> dict[str "usually non-normal null. `verdict` is the only claim the design supports: " "`distinguishable` when the test rejects, `equivalent_within_tolerance` when " "the gap from the null median is below the prespecified margin `delta`, and " - "`indeterminate` otherwise -- a failure to reject is never evidence of " - "equivalence. `sign_permute` is a **joint** target-set-and-sign null and is " + "`indeterminate` otherwise -- the point-gap rule is not a formal " + "equivalence test. `sign_permute` is a **joint** target-set-and-sign null and is " "off the ladder (`on_ladder = False`); the rank-3 rung is the weight-matched " "transmitter null, which holds each transmitter's share of total outgoing " "weight fixed (T23).", @@ -2091,10 +2114,9 @@ def arm(compound: str, assay: str, readout: str, n_i: int, tag: str) -> dict[str ) ctx.note( "instrument validation: %s Detection over the planted grid by permutation " - "count is %s and by planted strength %s; the empirical false-positive rate on " - "the unplanted control is %d of %d. Effect size dominates the permutation " - "count: full power at planted strength %s, detection at or below one in four " - "at %s." + "count is %s and by planted strength %s; unplanted controls were detected in " + "%d of %d runs. In this finite grid, observed detection was 1.00 at planted " + "strength %s and at most 0.25 at %s; these fractions are not general power estimates." % ( rec["statement"], ctx.text("val_power_by_n"), @@ -2150,14 +2172,16 @@ def arm(compound: str, assay: str, readout: str, n_i: int, tag: str) -> dict[str "p_weight", "p_degree", ], - "Ground-truth recovery and power for the dependence ladder. A recurrent " - "cholinergic cycle of known strength is planted in a synthetic cut of the same " - "size, density and transmitter composition as the `named` cut, and a gain patch " + "Limited planted-cycle recovery check for the dependence ladder. A recurrent " + f"cholinergic cycle of known strength is planted in a synthetic {ctx.get('val_recovery_nodes')}-node cut " + f"with about {ctx.get('val_recovery_edges')} background edges. Its transmitter mix was chosen close to " + "the `named` cut, but size and connectivity were not matched. A gain patch " "that collapses `g_ach` removes an amplification that exists only while the " "cycle is intact -- so the drug *contrast*, not merely the rate, depends on the " - "wiring. `loop_strength = 0` plants nothing and is the negative control, whose " - "detection rate is the empirical false-positive rate. Nothing here touches the " - "connectome or the compound library: the experiment is about the instrument.", + "wiring. `loop_strength = 0` plants nothing and is the negative control. " + "The small sample gives an observed control count, not a precise false-positive " + "rate or power estimate for the paper's endpoint. Nothing here touches the " + "connectome or the compound library.", md_fields=["experiment", "loop_strength", "n", "replicates", "detection_rate", "class"], ) @@ -2195,7 +2219,7 @@ def landscape(graph: str, n_i: int, tag: str) -> dict[str, Any]: f"dep_{tag}_below_relative_floor_cells", [f"{r['compound']}@{r['conc_M']:.0e}" for r in s["below_relative_effect_floor"]], text=", ".join( - f"{r['compound']} {_fmt_M(r['conc_M'])} ({r['real_effect']:.3g} Hz)" + f"{r['compound']} {_fmt_M(r['conc_M'])} ({r['real_effect']:.3g} rate-model units)" for r in s["below_relative_effect_floor"] ) or "none", @@ -2233,8 +2257,8 @@ def landscape(graph: str, n_i: int, tag: str) -> dict[str, Any]: for v, k in verdicts.items(): ctx.put(f"dep_{tag}_verdicts_{v}", int(k)) deltas = land.get("delta_range") or [None, None] - ctx.put(f"dep_{tag}_delta_min", _f(deltas[0]), unit="Hz") - ctx.put(f"dep_{tag}_delta_max", _f(deltas[1]), unit="Hz") + ctx.put(f"dep_{tag}_delta_min", _f(deltas[0]), unit="rate-model units") + ctx.put(f"dep_{tag}_delta_max", _f(deltas[1]), unit="rate-model units") # the cells the paper names one by one for cell in land["cells"]: if abs(cell["conc_M"] - PAPER_CONC) > 1e-18: @@ -2272,11 +2296,11 @@ def landscape(graph: str, n_i: int, tag: str) -> dict[str, Any]: ) if land_taste["summary"]["class_counts"].get("composition-dominated", 0) == 0: ctx.note( - "the central RQ2 result REVERSES on a cut that has topology: repeated on " + "the descriptive class distribution differs on a denser cut: repeated on " "`taste_motor` (%d nodes, %d edges, mean degree %.1f) the landscape returns " "%d composition-dominated cells of %d, against %d of %d on the in-star " "`named` cut. The composition-dominated majority is a property of the " - "substrate, not of receptor perturbation on connectomes." + "tested cut and runtime, not a general result about connectomes." % ( ctx.get("taste_nodes") or 0, ctx.get("taste_edges") or 0, @@ -2394,10 +2418,25 @@ def landscape_rows(res: dict[str, Any]) -> list[dict[str, Any]]: ) # ---- F6: the profile itself -------------------------------------- - fig, axes = plt.subplots(2, len(MODES), figsize=(12.0, 5.4)) + fig_modes = ( + "sign_permute_weight_matched", "weight_permute", + "rewire_degree_preserving", "erdos_renyi", + ) + mode_titles = { + "sign_permute_weight_matched": "Weight-matched labels", + "weight_permute": "Shuffled weights", + "rewire_degree_preserving": "Degree-preserving rewiring", + "erdos_renyi": "Random graph", + } + verdict_titles = { + "distinguishable": "distinguishable", + "equivalent_within_tolerance": "within point-gap tolerance", + "indeterminate": "not resolved", + } + fig, axes = plt.subplots(2, len(fig_modes), figsize=(12.0, 5.4)) for row, compound in enumerate(("imidacloprid", "fipronil")): sub = {r["mode"]: r for r in rows if r["assay"] == "subgraph" and r["compound"] == compound} - for col, mode in enumerate(MODES): + for col, mode in enumerate(fig_modes): ax = axes[row, col] vals = np.asarray(draws.get((compound, mode), []), dtype=float) real = sub[mode]["real_effect"] @@ -2408,11 +2447,9 @@ def landscape_rows(res: dict[str, Any]) -> list[dict[str, Any]]: p = sub[mode]["p_two_sided"] verdict = str(sub[mode].get("verdict") or "") ax.set_title( - f"{mode}\np = {'n/a' if p is None else f'{p:.3f}'}\n{verdict.replace('_', ' ')}", + f"{mode_titles[mode]}\np = {'n/a' if p is None else f'{p:.3f}'}\n{verdict_titles.get(verdict, verdict)}", fontsize=7.6, - color=CRIT if verdict == "distinguishable" else ( - GOOD if verdict == "equivalent_within_tolerance" else TEXT2 - ), + color=CRIT if verdict == "distinguishable" else TEXT2, ) ax.tick_params(labelsize=7) ax.grid(axis="y", lw=0.4) @@ -2420,7 +2457,17 @@ def landscape_rows(res: dict[str, Any]) -> list[dict[str, Any]]: if col == 0: ax.set_ylabel(f"{compound}\nshuffles") if row == 1: - ax.set_xlabel("effect on mean rate (Hz)") + ax.set_xlabel("effect on mean rate (model units)") + if col == len(fig_modes) - 1 and vals.size: + # The Erdős–Rényi null concentrates near zero. Show its shape + # without hiding the real-cut contrast on the main axis. + lo, hi = np.quantile(vals, [0.005, 0.995]) + if hi > lo: + inset = ax.inset_axes([0.49, 0.47, 0.47, 0.42]) + inset.hist(vals, bins=25, color=SEQ[1], edgecolor="white", linewidth=0.2) + inset.set_xlim(float(lo), float(hi)) + inset.tick_params(labelsize=5) + inset.set_title("null detail", fontsize=6) axes[0, 0].legend(fontsize=7) fig.suptitle( f"Connectome-dependence profile, {n} permutations per mode " @@ -2435,17 +2482,18 @@ def landscape_rows(res: dict[str, Any]) -> list[dict[str, Any]]: "Histograms are the drug effect (treated minus vehicle on the network mean " f"rate) measured on {n} degraded copies of the 1-hop MN9/DNp01 `named` cut; the " "red line is the same contrast on the real cut. Columns run in increasing order " - "of destruction: transmitter labels permuted, synapse weights permuted, " - "degree-preserving double-edge rewiring, Erdos-Renyi. Titles carry the " + "of destruction: transmitter labels permuted while holding outgoing-weight " + "shares near their real values, synapse weights permuted, " + "degree-preserving double-edge rewiring, Erdős–Rényi. Insets show the " + "central 99% of the concentrated Erdős–Rényi null distributions. Titles carry the " "empirical two-sided permutation p and the three-way verdict against the " - f"prespecified equivalence margin (delta = " - f"{ctx.text('dep_named_imidacloprid_delta')} Hz, 5 % of the vehicle readout): " - "`distinguishable` means the test rejected, `equivalent within tolerance` means " - "the gap from the null median is below the margin, and `indeterminate` means " - "neither -- a failure to reject is not evidence that the degraded graph gives " - "the same effect. Plain label permutation is shown because it is informative, " - "but it moves the weighted excitation/inhibition balance as well as transmitter " - "identity (T23), so it is a joint null and not a rung of the ladder. " + f"declared point-gap margin (delta = " + f"{ctx.text('dep_named_imidacloprid_delta')} model rate units, 5 % of the vehicle readout): " + "`distinguishable` means the test rejected, `within point-gap tolerance` " + "means the gap from the null median is below the margin after non-rejection, " + "and `not resolved` means neither -- no label is a statistical equivalence " + "test. The plain label permutation is documented in T23: it moves weighted " + "excitation/inhibition balance and is a joint null rather than a ladder rung. " f"Imidacloprid is {imi['class']} (necessary level: " f"{imi['necessary_information_level'].get('level')}, " f"{imi['necessary_information_level'].get('verdict')}); fipronil is " @@ -2522,15 +2570,17 @@ def landscape_rows(res: dict[str, Any]) -> list[dict[str, Any]]: "Classes are decided on Benjamini-Hochberg adjusted probabilities across each " "landscape's structural tests, with a prespecified relative effect floor of 1 % " f"of the vehicle readout. Left, the 1-hop `named` cut: " - f"{counts.get('topology-dependent', 0)} of {land['n_cells']} cells need the " - f"wiring pattern, {counts.get('composition-dominated', 0)} are not " - "distinguishable from any structure-preserving degradation, and " - f"{counts.get('no-effect', 0)} do not move the readout at all. Right, the " + f"{counts.get('topology-dependent', 0)} of {land['n_cells']} cells receive the " + f"topology-dependent label, {counts.get('composition-dominated', 0)} receive " + "composition-dominated and " + f"{counts.get('no-effect', 0)} fall below the effect floor. Right, the " f"denser `taste_motor` cut: {ct.get('topology-dependent', 0)} topology-dependent, " f"{ct.get('composition-dominated', 0)} composition-dominated, " f"{ct.get('no-effect', 0)} no-effect. The composition-dominated majority does " "not survive the change of substrate, and the `named` cut is an in-star whose " - "degree-preserving rewire is close to the identity (T19).", + "degree-preserving rewire is close to the identity (T19). These are " + "descriptive classes formed from BH-adjusted component tests, without a " + "separate cell-level FDR guarantee.", ) # ---- F16: the instrument's own power ------------------------------ @@ -2560,7 +2610,7 @@ def landscape_rows(res: dict[str, Any]) -> list[dict[str, Any]]: ax2.set_xticks(range(len(labels))) ax2.set_xticklabels(labels) ax2.set_xlabel("planted loop strength") - ax2.set_ylabel("|drug effect| on the synthetic cut (Hz)") + ax2.set_ylabel("|drug effect| on the synthetic cut (rate-model units)") ax2.set_title( f"Ground-truth recovery at n = {rec['n']}\n" f"{rec['n_detected']} of {rec['n_positive']} planted effects recovered, " @@ -2576,17 +2626,19 @@ def landscape_rows(res: dict[str, Any]) -> list[dict[str, Any]]: fig, "F16_instrument_validation", "The dependence ladder pointed at a graph whose answer is known in advance. A " - "recurrent cholinergic cycle of known strength is planted in a synthetic cut of " - "the same size, density and transmitter composition as the `named` cut; a gain " + "recurrent cholinergic cycle of known strength is planted in a synthetic " + f"{ctx.get('val_recovery_nodes')}-node cut with about " + f"{ctx.get('val_recovery_edges')} background edges; it is not matched to " + "the real extract's size or connectivity. A gain " "patch that collapses `g_ach` then removes an amplification that exists only " "while the cycle is intact, so the drug contrast depends on the wiring by " "construction. Left: the share of independently generated graphs the ladder " "classifies topology-dependent, against planted strength and permutation count. " "Right: the recovery experiment itself, with red marking the strengths the " - "ladder recovered. The unplanted control is the empirical false-positive rate " + "ladder recovered. The unplanted control yielded a detection count " f"({ctx.get('val_power_control_hits')} of " - f"{ctx.get('val_power_control_runs')}). Without this panel, a negative from the " - "ladder would be indistinguishable from an under-powered test.", + f"{ctx.get('val_power_control_runs')}). This is a limited positive-control " + "check, not a general power estimate for the paper's biological endpoints.", ) def step_landscape(ctx: Ctx) -> None: @@ -2741,8 +2793,8 @@ def step_dose(ctx: Ctx) -> None: ctx.put("ic50_hi_M", float(ci["ic50"][1]), text=_fmt_M(ci["ic50"][1])) ctx.put("ic50_slope", float(fit["slope"])) ctx.put("ic50_r2", float(fit["r2"]), text=f"{fit['r2']:.4f}") - ctx.put("ic50_top_hz", float(fit["top"]), unit="Hz") - ctx.put("ic50_bottom_hz", float(fit["bottom"]), unit="Hz") + ctx.put("ic50_top_hz", float(fit["top"]), unit="rate-model units") + ctx.put("ic50_bottom_hz", float(fit["bottom"]), unit="rate-model units") ctx.put("ic50_in_range", bool(fit["in_range"])) ctx.put("ic50_readout", ic["readout"]) ctx.put( @@ -2754,14 +2806,14 @@ def step_dose(ctx: Ctx) -> None: tor = sensitivity("subgraph", "imidacloprid", PAPER_CONC, readout="mean_hz", seed=ctx.seed) tor = sorted(tor, key=lambda r: abs(float(r["span"])), reverse=True) for r in tor: - ctx.put(f"tornado_span_{r['param']}", float(r["span"]), unit="Hz") + ctx.put(f"tornado_span_{r['param']}", float(r["span"]), unit="rate-model units") ctx.put("tornado_top_param", tor[0]["param"]) zero_span = [r["param"] for r in tor if abs(float(r["span"])) < 1e-9] ctx.put("tornado_zero_span_params", zero_span, text=", ".join(zero_span) or "none") if "ec50" in zero_span: ctx.note( "at 1 uM imidacloprid the model is exactly insensitive to a 2x change in " - "the insect_nAChR EC50 and in the Hill coefficient (span 0 Hz): occupancy " + "the insect_nAChR EC50 and in the Hill coefficient (span 0 rate-model units): occupancy " "is saturated, so the gain-rule coefficient is the only parameter that " "still moves the readout. A dose-response experiment, not a single high " "dose, is what constrains the EC50." @@ -2780,7 +2832,7 @@ def step_dose(ctx: Ctx) -> None: ax.axvline(fit["ic50"], color=CRIT, lw=1, ls="--") ax.set_xscale("log") ax.set_xlabel("imidacloprid concentration (M)") - ax.set_ylabel(f"network {ic['readout']} (Hz)") + ax.set_ylabel(f"network {ic['readout']} (rate-model units)") ax.set_title( f"Model IC50 = {_fmt_M(fit['ic50'])}\n" f"[{_fmt_M(ci['ic50'][0])}, {_fmt_M(ci['ic50'][1])}], slope {fit['slope']:.2f}, r2 {fit['r2']:.4f}" @@ -2800,7 +2852,7 @@ def step_dose(ctx: Ctx) -> None: ax2.axvline(base, color=TEXT1, lw=1) ax2.set_yticks(y) ax2.set_yticklabels(names) - ax2.set_xlabel(f"network mean rate (Hz); treated baseline at 1 uM = {base:.2f} Hz") + ax2.set_xlabel(f"network mean output (rate-model units); treated baseline at 1 uM = {base:.2f}") ax2.set_title("One-at-a-time sensitivity\n(each parameter halved and doubled)") ax2.grid(axis="x", lw=0.5) ax2.set_axisbelow(True) @@ -2931,7 +2983,7 @@ def step_genotype(ctx: Ctx) -> None: ax.set_xticks(x) ax.set_xticklabels(labels, fontsize=7, rotation=20, ha="right") ax.set_ylim(0, 1.15) - ax.set_ylabel("target occupancy at 1 uM") + ax.set_ylabel("target engagement proxy at 1 uM") ax.set_title(title, fontsize=9.5) ax.grid(axis="y", lw=0.5) ax.set_axisbelow(True) @@ -2941,7 +2993,7 @@ def step_genotype(ctx: Ctx) -> None: ctx, fig, "F9_genotype_panel", - "Target-receptor occupancy at 1 uM for wild type and each published resistance " + "Model target engagement at 1 uM for wild type and each published resistance " "allele that has a sourced fold-shift for that compound. Grey bars are alleles " "with no sourced number for the compound: they are left identical to wild type " "and reported, never extrapolated. Left and centre: fipronil and imidacloprid. " @@ -3376,7 +3428,7 @@ def step_expression(ctx: Ctx) -> None: from flylab.pharm.expression import coverage, expression_table, run_weighted_subgraph_assay cov = coverage() - ctx.put("expr_coverage_overall", float(cov["fraction_known_overall"])) + ctx.put("expr_coverage_overall", float(cov["fraction_cell_receptor_pairs_annotated"])) for receptor, d in cov["graphs"]["named"].items(): if isinstance(d, dict) and "fraction_known" in d: ctx.put(f"expr_cov_named_{receptor}", float(d["fraction_known"])) @@ -3391,11 +3443,11 @@ def step_expression(ctx: Ctx) -> None: nb = run_weighted_subgraph_assay("imidacloprid", PAPER_CONC) r = nb["readouts"] - ctx.put("expr_mn9_uniform_hz", float(r["uniform"]["mn9_hz"]), unit="Hz") - ctx.put("expr_mn9_weighted_hz", float(r["mn9_hz"]), unit="Hz") - ctx.put("expr_mn9_delta_hz", float(r["delta_vs_uniform"]["mn9_hz"]), unit="Hz") - ctx.put("expr_mean_uniform_hz", float(r["uniform"]["mean_hz"]), unit="Hz") - ctx.put("expr_mean_weighted_hz", float(r["mean_hz"]), unit="Hz") + ctx.put("expr_mn9_uniform_hz", float(r["uniform"]["mn9_hz"]), unit="rate-model units") + ctx.put("expr_mn9_weighted_hz", float(r["mn9_hz"]), unit="rate-model units") + ctx.put("expr_mn9_delta_hz", float(r["delta_vs_uniform"]["mn9_hz"]), unit="rate-model units") + ctx.put("expr_mean_uniform_hz", float(r["uniform"]["mean_hz"]), unit="rate-model units") + ctx.put("expr_mean_weighted_hz", float(r["mean_hz"]), unit="rate-model units") def step_ablation(ctx: Ctx) -> None: @@ -3492,7 +3544,7 @@ def step_ablation(ctx: Ctx) -> None: paper = tab["information_gain"][f"{1e-6:.3e}"] by_name = {lvl["level"]: lvl for lvl in paper["levels"]} ctx.put("abl_rho_topology_paper", _f(by_name["C_topology_only"].get("spearman_rho_vs_full"))) - ctx.put("abl_rms_topology_hz", _f(by_name["C_topology_only"].get("residual_rms_hz")), unit="Hz") + ctx.put("abl_rms_topology_hz", _f(by_name["C_topology_only"].get("residual_rms_hz")), unit="rate-model units") ctx.put("abl_rho_composition_paper", _f(by_name["B_composition_only"].get("spearman_rho_vs_full"))) ctx.put("abl_rho_receptor_paper", _f(by_name["A_receptor_only"].get("spearman_rho_vs_full"))) paper_cmp = paper.get("generic_rule_comparison") or {} @@ -3548,7 +3600,7 @@ def step_ablation(ctx: Ctx) -> None: if (ctx.get("abl_rho_topology_paper") or 0) >= (ctx.get("abl_rho_topology_floor_paper") or 0) + 0.3: ctx.note( - "the level-C conclusion is withdrawn: with a direction-aware generic rule " + "the level-C rank comparison depends on the generic rule: with a direction-aware rule " "the topology-only baseline's rank correlation with the full model at 1 uM " "rises from %s (depression-only floor, which cannot express disinhibition " "and whose every entry is at most zero) to %s. 'The connectome without the " @@ -3587,7 +3639,7 @@ def step_ablation(ctx: Ctx) -> None: if cond["shuffled_compound_assignment"]["identical_to_observed"]: ctx.note( "the composition-versus-full correlation carries no compound-level " - "information: shuffling the compound-to-gain assignment leaves it " + "evidence from a label shuffle: shuffling the compound-to-gain assignment leaves it " "identically %s, and pharmacology-free pseudo-compounds whose gain vectors " "have the shape the mechanism rules produce already reach a median of %s. " "Both levels are functions of the same gain vector, so the number is a " @@ -3638,11 +3690,11 @@ def step_ablation(ctx: Ctx) -> None: "Reference distribution for the composition-versus-full rank correlation. The " "composition level and the full model are not independent models: both are " "functions of the same gain vector, so the correlation has a large structural " - "floor and a bare value near 0.99 is not interpretable. The matched reference " + "floor and the observed value must be compared with a matched reference. The latter " "draws pseudo-compounds with the shape the shipped mechanism rules produce " "(one receptor, one transmitter, one gain moved) and no pharmacology at all; " "shuffling the compound labels leaves the observed value unchanged, which is a " - "proof rather than a coincidence.", + "consequence of moving the paired B and D values together.", md_fields=["condition", "spearman_rho", "n_compounds", "draws"], ) @@ -3686,7 +3738,7 @@ def step_ablation(ctx: Ctx) -> None: "the composition verdict is not robust to the engine's row normalisation: " "the B-versus-D rank correlation at 1 uM is %s under the shipped `row_abs` " "mode and %s under a degree-corrected one. The normalisation is an " - "undocumented modelling choice that makes each cell's recurrent input a " + "documented modelling choice that makes each cell's recurrent input a " "composition-weighted average of its presynaptic gains." % ( ctx.text("abl_rho_composition_row_abs"), @@ -3748,8 +3800,8 @@ def step_ablation(ctx: Ctx) -> None: "compound-concentration cell from a model that has been denied one layer of " "information: receptor engagement only (dimensionless), mechanism gains on the " "graph's transmitter composition only (excitation index), the real cut with one " - "generic multiplier instead of mechanism-specific gains (Hz), and the full " - "model (Hz). The generic multiplier is direction-aware: generic in magnitude, " + "generic multiplier instead of mechanism-specific gains (rate-model units), " + "and the full model (rate-model units). The generic multiplier is direction-aware: generic in magnitude, " "with one bit of sign taken from the mechanism table, which is the minimum a " "connectome-without-pharmacology model needs to order a library containing " "disinhibitors. `C_topology_only_floor` is the historical depression-only rule, " @@ -3780,9 +3832,9 @@ def step_ablation(ctx: Ctx) -> None: "have credited it.", ) - fig, ax = plt.subplots(figsize=(7.4, 4.0)) + fig, ax = plt.subplots(figsize=(8.2, 4.1)) xs = list(range(len(LEVELS))) - for i, conc_key in enumerate(sorted(tab["information_gain"])): + for i, conc_key in enumerate(sorted(tab["information_gain"], key=float)): gain = tab["information_gain"][conc_key] by = {lvl["level"]: lvl for lvl in gain["levels"]} ys = [by[lvl].get("spearman_rho_vs_full") for lvl in LEVELS] @@ -3797,19 +3849,27 @@ def step_ablation(ctx: Ctx) -> None: ax.plot( [xs[LEVELS.index("C_topology_only")]], [floor_rho], "x", color=SEQ[min(i + 1, 4)], ms=9, mew=2, + label="depression-only comparator" if i == 0 else None, ) ax.axhline(0.0, color=TEXT2, lw=1) - ax.axhline(REPRODUCES_RHO, color=GOOD, lw=1, ls="--") - ax.text( - 0.02, REPRODUCES_RHO + 0.02, "reproduces the full ordering", fontsize=7, - color=GOOD, transform=ax.get_yaxis_transform(), - ) + x_b = xs[LEVELS.index("B_composition_only")] + if sparse.get("p05") is not None and sparse.get("p95") is not None: + ax.fill_betweenx( + [float(sparse["p05"]), float(sparse["p95"])], + x_b - 0.10, x_b + 0.10, color=GOOD, alpha=0.22, + label="B versus D reference 5th–95th percentile", + ) + if sparse.get("median") is not None: + ax.plot( + [x_b - 0.11, x_b + 0.11], [float(sparse["median"])] * 2, + color=GOOD, lw=1.5, ls="--", label="B versus D reference median", + ) ax.set_xticks(xs) - ax.set_xticklabels([short.get(lvl, lvl).replace("_", " ") for lvl in LEVELS]) - ax.set_ylabel("signed Spearman rho of the library ordering vs the full model") - ax.set_xlabel("information the level is allowed") + ax.set_xticklabels([lvl.split("_", 1)[0] for lvl in LEVELS]) + ax.set_ylabel("Signed Spearman ρ vs full model") + ax.set_xlabel("Ablation level") ax.set_title(f"Ablation ladder over the {len(tab['compounds'])}-compound library") - ax.legend(fontsize=7.5, title="concentration", title_fontsize=7.5) + ax.legend(fontsize=7.2, loc="center left", bbox_to_anchor=(1.01, 0.5)) ax.grid(lw=0.5) ax.set_axisbelow(True) fig.tight_layout() @@ -3818,18 +3878,20 @@ def step_ablation(ctx: Ctx) -> None: fig, "F13_ablation", f"Each level predicts the same {len(tab['compounds'])} compounds with one layer " - "of information removed, and is scored by how well it reproduces the full " + "of information removed, and is scored by rank agreement with the full " "model's *ordering* of the library (levels have different units, so values are " "not comparable; the correlation is signed, so a perfect inversion scores -1 " "rather than 1). Crosses mark the historical depression-only level-C rule, " "which applied a multiplier of at most 1.0 to every transmitter alike and so " - "could not express disinhibition: its poor correlation was structural, not a " - "finding about connectomes, and the conclusion drawn from it is withdrawn. " + "could not express disinhibition: its poor correlation follows partly from " + "that restriction. " f"With a direction-aware rule level C rises to " f"{ctx.text('abl_rho_topology_paper')} at 1 uM from " - f"{ctx.text('abl_rho_topology_floor_paper')}. The composition level's agreement " - "with the full model must be read against the matched reference distribution " - "of T24, not on its own.", + f"{ctx.text('abl_rho_topology_floor_paper')}. The short green interval at B " + "shows the 5th–95th percentile of the matched one-gain-per-pseudo-compound " + "B-versus-D reference distribution (T24), with its median dashed. The " + "composition level's " + "agreement with the full model must be read against that distribution.", ) def step_stability(ctx: Ctx) -> None: @@ -3890,8 +3952,8 @@ def step_stability(ctx: Ctx) -> None: "the C3 retention rate is no longer vacuous and is no longer uniform: at " "n_shuffles = %d (resolution %.4f, against the previous 6 at which p <= " "0.05 was arithmetically unattainable) the conclusion is retained by %d of " - "%d specifications, but only %d of those reach equivalence within the " - "prespecified margin; %d are merely indeterminate." + "%d specifications, but only %d of those pass the descriptive point-gap " + "rule; %d are indeterminate." % ( int(res["n_shuffles"]), float(res["shuffle_resolution"]), @@ -3925,8 +3987,8 @@ def step_stability(ctx: Ctx) -> None: with mechanism_spec(spec): nb = run_subgraph_assay("imidacloprid", PAPER_CONC) r = nb["readouts"] - ctx.put(f"stab_{label}_treated_hz", _f(r.get("mean_hz")), unit="Hz") - ctx.put(f"stab_{label}_vehicle_hz", _f((r.get("vehicle") or {}).get("mean_hz")), unit="Hz") + ctx.put(f"stab_{label}_treated_hz", _f(r.get("mean_hz")), unit="rate-model units") + ctx.put(f"stab_{label}_vehicle_hz", _f((r.get("vehicle") or {}).get("mean_hz")), unit="rate-model units") ctx.put(f"stab_{label}_g_ach", _f((nb.get("gains") or {}).get("g_ach"))) t, v = ctx.get("stab_monotone_treated_hz"), ctx.get("stab_monotone_vehicle_hz") if t and v: @@ -3934,7 +3996,7 @@ def step_stability(ctx: Ctx) -> None: ctx.note( "the suppression conclusion is specification-dependent: %d of %d " "prespecified specifications reverse it. Under %s the same compound at the " - "same engagement EXCITES the network (%s Hz treated vs %s Hz vehicle) " + "same engagement raises the model readout (%s treated vs %s vehicle, rate-model units) " "instead of suppressing it. Suppression is a property of the biphasic " "desensitisation term, not of the pharmacology-connectome integration." % ( @@ -4240,21 +4302,20 @@ def step_stability(ctx: Ctx) -> None: ctx, fig, "F14_conclusion_stability", - "Each row is a conclusion this paper could state; each column is one " - "prespecified, admissible way of turning receptor engagement into synaptic " - "gain. Green retains the conclusion, red reverses it, amber means the readout " + "Each row is a model-output predicate; each column is one tested way of " + "turning receptor engagement into synaptic gain, with transmitter signs " + "and synapse assignment fixed. Green retains the predicate, red reverses it, " + "and amber means the readout " "does not exist under that specification. The topology rows are decided on " f"{res['n_shuffles']} permutations per specification and on Benjamini-Hochberg " f"adjusted probabilities across the {res['n_structural_tests']} structural " - "tests of the run; until v0.6.1 the specification never reached the engine " - "those rows were computed on, so their uniformity was guaranteed by " - "construction rather than measured. Retained by every specification: " + "component tests of the run; this is not a cell-level FDR guarantee. " + "Retained by every tested specification: " + (", ".join(r["conclusion"] for r in res["rows"] if not r["fragile"]) or "none") - + ". The nicotinic suppression conclusion is reversed by every monotone rule, " - "which is how a result that depends on a modelling choice looks when it is " - "tested rather than asserted; and for a conclusion shaped as a failure to " - "reject, the fraction retained is only as strong as the equivalence behind it " - "(T14's `n_equivalent_within_tolerance` against `n_indeterminate`).", + + ". The nicotinic suppression predicate reverses under every monotone rule " + "in this family. For a predicate based on non-rejection, T14 separates " + "descriptive point-gap passes from indeterminate cases; neither is a formal " + "equivalence result.", ) @@ -4280,8 +4341,8 @@ def step_uncertainty(ctx: Ctx) -> None: ctx.put("unc_n_factors", int(res["n_factors"])) ctx.put("unc_evaluations", int(res["n_evaluations"])) ctx.put("unc_estimator", str(res["estimator"])) - ctx.put("unc_output_mean", _f(res["output_mean"]), unit="Hz") - ctx.put("unc_variance", _f(res["output_variance"]), unit="Hz^2") + ctx.put("unc_output_mean", _f(res["output_mean"]), unit="rate-model units") + ctx.put("unc_variance", _f(res["output_variance"]), unit="rate-model units^2") ctx.put("unc_interaction_share", _f(res["interaction_share"])) ctx.put("unc_interaction_share_clipped", _f(res["interaction_share_clipped"])) ctx.put("unc_sum_first_order", _f(res["sum_first_order"])) @@ -4391,10 +4452,10 @@ def step_uncertainty(ctx: Ctx) -> None: ) v = voi(result=res) - ctx.put("voi_var_total", _f(v["output_variance"]), unit="Hz^2") + ctx.put("voi_var_total", _f(v["output_variance"]), unit="rate-model units^2") for i, r in enumerate(v["rows"][:3], 1): ctx.put(f"voi_rank{i}_factor", r["factor"]) - ctx.put(f"voi_rank{i}_var", _f(r["voi_var"]), unit="Hz^2") + ctx.put(f"voi_rank{i}_var", _f(r["voi_var"]), unit="rate-model units^2") ctx.put(f"voi_rank{i}_fraction", _f(r["voi_fraction"])) ctx.put(f"voi_rank{i}_experiment", r["experiment"]) ctx.put(f"voi_rank{i}_cost", r["cost"]) @@ -4448,17 +4509,18 @@ def step_uncertainty(ctx: Ctx) -> None: "cost", "blocking_gate", ], - "Value of information. VOI_j = S_j x Var(Y) is the model variance that would " - "disappear if assumption j were resolved exactly while everything else stayed " - "as uncertain as it is; the upper bound uses the total-order index and is what " - "resolving j *last* would buy. A value of information cannot be negative, so a " + "Author-defined variance-priority heuristic. H_j = max(0, S_j) × Var(Y) " + "ranks variance associated with factor j under assumed input ranges; it is " + "not a formal expected value of information. The upper-bound column uses " + "the total-order index. The heuristic cannot be negative, so a " "negative first-order estimate is clipped to zero for the decision value and " "kept unclipped in `voi_fraction_raw`. `state` says whether the sample could " "resolve the factor at all: an unresolved factor carries no ranking claim, " "which is a statement about the sample size rather than about the factor. This " - "is variance of a model output under assumed input ranges, not an expected gain " - "in accuracy about a living fly.", + "uses squared rate-model units, despite legacy column names ending in hz2; " + "it is not an expected gain in accuracy about a living fly.", md_fields=["rank", "factor", "state", "voi_fraction_of_var", "voi_var_hz2", "experiment", "cost"], + lineterminator="\n", ) import numpy as np @@ -4507,8 +4569,8 @@ def step_uncertainty(ctx: Ctx) -> None: ax2.barh(y2, [r["voi_var"] for r in top], color=[GOOD if r.get("cost") == "none (compute only)" else SEQ[2] for r in top]) ax2.set_yticks(y2) ax2.set_yticklabels([r["factor"] for r in top], fontsize=8) - ax2.set_xlabel("model variance an exact answer would remove (Hz$^2$)") - ax2.set_title(f"Value of information (Var(Y) = {res['output_variance']:.3g} Hz$^2$)", fontsize=10) + ax2.set_xlabel("variance-priority score (rate-model units$^2$)") + ax2.set_title(f"Priority heuristic (Var(Y) = {res['output_variance']:.3g} model units$^2$)", fontsize=10) ax2.grid(axis="x", lw=0.5) ax2.set_axisbelow(True) fig.tight_layout() @@ -4529,7 +4591,7 @@ def step_uncertainty(ctx: Ctx) -> None: f"against lif_seed at {ctx.text('unc_null_factor_S')}). Right: the same shares " "scaled back into the readout's variance and mapped onto the experiment that " "would resolve each assumption; a negative estimate is clipped to zero for the " - "decision value, because a value of information cannot be negative. Green marks " + "heuristic score. This is not formal expected value of information. Green marks " "a factor that needs no experiment at all, only a re-analysis of data already " "held.", ) @@ -4615,6 +4677,7 @@ def step_claims(ctx: Ctx) -> None: "FlyLab; `COMPUTED` is arithmetic on the links above it. The same function " "splits the result into facts, model inference and unknowns.", md_fields=["step", "label", "classification", "contributes"], + lineterminator="\n", ) @@ -4743,9 +4806,9 @@ def step_scale(ctx: Ctx) -> None: # --- the paper-quotable shape of the result ------------------- # Rows, smallest cut first. The key methodological point is that the - # inversion does NOT track node count: `scale_1k` has fewer nodes than - # `named` and a mean degree twenty times higher, and it already comes - # out topology-dependent. What separates them is recurrence. + # The saved extracts associate the verdict change more with recurrence + # than node count, but their differing construction and structure do not + # isolate recurrence as a cause. rows_by_size = sorted(study["rows"], key=lambda r: (int(r["n_nodes"]), int(r["n_edges"]))) ns = [int(r["n"]) for r in rows_by_size] ctx.put("scale_study_n_range", [min(ns), max(ns)], text=f"{min(ns)}-{max(ns)}") @@ -4754,11 +4817,17 @@ def step_scale(ctx: Ctx) -> None: comp_cuts = sorted({r["cut"] for r in rows_by_size if r.get("class") == "composition-dominated"}) ctx.put("scale_study_composition_dominated_cuts", comp_cuts, text=", ".join(comp_cuts) or "none") ctx.put("scale_study_n_composition_dominated", len(comp_cuts)) - every_null = sorted( + real_connectome_ladder_cuts = sorted( {r["cut"] for r in rows_by_size if r.get("necessary_information_level") == "real_connectome"}, key=lambda c: next(int(x["n_nodes"]) for x in rows_by_size if x["cut"] == c), ) - ctx.put("scale_study_every_null_cuts", every_null, text=", ".join(every_null) or "none") + # This necessary-information ladder ranks the three edge/wiring modes; + # transmitter-label shuffles are orthogonal checks and are not included. + ctx.put( + "scale_study_every_null_cuts", + real_connectome_ladder_cuts, + text=", ".join(real_connectome_ladder_cuts) or "none", + ) # Rungs whose permutation budget is so small that the smallest # attainable probability is within a factor of two of alpha: a # rejection there is the smallest the test can express, not a strong @@ -5063,7 +5132,6 @@ def sub(m: re.Match[str]) -> str: #: none has rotted into a wish list; regenerate this tuple from the two #: templates whenever their prose changes. PAPER_KEYS: tuple[str, ...] = ( - "abl_composition_survives_normalisations", "abl_compounds", "abl_concs", "abl_generic_floor_note", @@ -5099,12 +5167,9 @@ def sub(m: re.Match[str]) -> str: "bal_tol", "claims_chain_links", "claims_computed", - "claims_facts_phrase", - "claims_inference_phrase", "claims_literature_derived", "claims_model_assumption", "claims_observed_phrase", - "claims_unknown_phrase", "dep_land_below_relative_floor", "dep_land_below_relative_floor_cells", "dep_land_cells", @@ -5126,9 +5191,6 @@ def sub(m: re.Match[str]) -> str: "dep_land_non_monotone_cells", "dep_land_taste_cells", "dep_land_taste_composition_dominated", - "dep_land_taste_imidacloprid_class", - "dep_land_taste_imidacloprid_p_rewire_degree_preserving", - "dep_land_taste_imidacloprid_q", "dep_land_taste_n_structural_tests", "dep_land_taste_no_effect", "dep_land_taste_non_monotone", @@ -5276,7 +5338,6 @@ def sub(m: re.Match[str]) -> str: "rank_shared_source_verb", "rank_skipped", "rank_source_disjoint", - "rank_source_disjoint_be", "scale_1k_edges", "scale_1k_in_star", "scale_1k_mean_degree", diff --git a/tests/test_analysis_baselines.py b/tests/test_analysis_baselines.py index 6e566a4..f24db09 100644 --- a/tests/test_analysis_baselines.py +++ b/tests/test_analysis_baselines.py @@ -139,7 +139,7 @@ def test_composition_only_reproduces_the_full_model_for_nicotinic_agonists(): b = rows["B_composition_only"] assert b["spearman_rho_vs_full"] >= 0.9 assert b["reproduces_full_ordering"] is True - assert any("B_composition_only REPRODUCES" in s for s in out["statements"]) + assert any("B_composition_only meets the declared rank-agreement threshold" in s for s in out["statements"]) # and it is the composition layer specifically: the receptor scorecard on # its own gets the ordering wrong (it has no gain rule). assert rows["A_receptor_only"]["reproduces_full_ordering"] is False @@ -151,7 +151,7 @@ def test_receptor_only_does_not_reproduce_the_full_model(): out = ablation("imidacloprid", PAPER_CONC, compounds=SET) rows = {r["level"]: r for r in out["information_gain"]["levels"]} assert rows["A_receptor_only"]["reproduces_full"] is False - assert any("A_receptor_only does NOT reproduce" in s for s in out["statements"]) + assert any("A_receptor_only is below the declared rank threshold" in s for s in out["statements"]) def test_information_gain_can_be_switched_off(): @@ -309,7 +309,7 @@ def test_composition_reference_distribution_gives_an_unsurprising_range(): # shuffling compound labels cannot move the correlation: both levels are # functions of the same gain vector. That is the point of the condition. assert cond["shuffled_compound_assignment"]["identical_to_observed"] is True - assert any("EVIDENCE AGAINST" in s for s in out["statements"]) + assert any("observed value should be read against the matched reference" in s for s in out["statements"]) assert any("floor" in s for s in out["statements"]) json.dumps(out) diff --git a/tests/test_analysis_claims.py b/tests/test_analysis_claims.py index 2fc83de..9b90b14 100644 --- a/tests/test_analysis_claims.py +++ b/tests/test_analysis_claims.py @@ -111,8 +111,11 @@ def test_transmitters_are_flagged_as_predictions(audit): def test_expression_link_carries_the_coverage_and_the_mn9_gap(audit): expr = next(link for link in audit["chain"] if link["step"] == "expression_assumption") - fraction = expr["detail"]["fraction_cells_mapped"] + fraction = expr["detail"]["fraction_cell_receptor_pairs_annotated"] assert fraction is None or 0.0 < fraction < 1.0 + assert expr["detail"]["coverage_unit"] == "graph_node_x_receptor_key_pair" + assert "graph-node × receptor-key pairs" in expr["statement"] + assert "per-cell expression measurement" in expr["statement"] assert any("MN9" in u for u in expr["unknowns"]) @@ -280,4 +283,3 @@ def test_an_unsupported_row_reports_no_engagement(audit): if not row["modelled"]: assert row["engagement"] is None # N/A, never a small number assert row["param_value_M"] is None - diff --git a/tests/test_browser_smoke.py b/tests/test_browser_smoke.py new file mode 100644 index 0000000..27e44d6 --- /dev/null +++ b/tests/test_browser_smoke.py @@ -0,0 +1,185 @@ +"""A real-browser smoke test for the static Pyodide bench. + +This is intentionally opt-in: it builds a site, starts a local HTTP server, +downloads the pinned Pyodide runtime in Chromium, and runs one dashboard call. +""" +from __future__ import annotations + +import functools +import importlib.util +import json +import sys +import threading +from http.server import SimpleHTTPRequestHandler, ThreadingHTTPServer +from pathlib import Path + +import pytest + +REPO = Path(__file__).resolve().parents[1] + + +def _load_build_pages(): + spec = importlib.util.spec_from_file_location("build_pages", REPO / "scripts" / "build_pages.py") + module = importlib.util.module_from_spec(spec) + sys.modules["build_pages"] = module + assert spec.loader is not None + spec.loader.exec_module(module) + return module + + +class _QuietHandler(SimpleHTTPRequestHandler): + def log_message(self, _format, *args): + pass + + +@pytest.mark.slow +def test_static_bench_starts_runs_dashboard_and_falls_back_without_charts(tmp_path): + playwright_api = pytest.importorskip( + "playwright.sync_api", + reason="install the browser-test extra and Chromium to run the static-browser smoke test", + ) + + site = tmp_path / "pages" + _load_build_pages().build(out=site) + + handler = functools.partial(_QuietHandler, directory=str(site)) + server = ThreadingHTTPServer(("127.0.0.1", 0), handler) + server_thread = threading.Thread(target=server.serve_forever, daemon=True) + server_thread.start() + url = f"http://127.0.0.1:{server.server_port}/" + + try: + with playwright_api.sync_playwright() as playwright: + browser = playwright.chromium.launch(headless=True) + try: + page = browser.new_page() + page.set_default_timeout(30_000) + + def block_chart_assets(route): + request = route.request + path = request.url.lower() + is_chart_script = request.resource_type == "script" and any( + name in path for name in ("plotly", "cytoscape") + ) + if is_chart_script: + route.abort() + else: + route.continue_() + + page.route("**/*", block_chart_assets) + page.goto(url, wait_until="domcontentloaded", timeout=60_000) + + # The overlay remains visible on boot failure. Wait until the + # loader removes it, or surfaces its explicit error state. + page.wait_for_function( + "() => !document.querySelector('#flylab-boot') || " + "document.querySelector('#flylab-boot.fb-error') !== null", + timeout=10 * 60_000, + ) + boot_error = page.locator("#flylab-boot.fb-error .fb-label").count() + assert boot_error == 0, page.locator("#flylab-boot").inner_text() + assert page.evaluate("window.FLYLAB_STATIC") is True + + # The first dashboard activation is automatic. Wait for its + # data and verify the chart-less path leaves the equivalent table. + page.wait_for_function( + "() => document.querySelectorAll('#dash-tiles .tile').length >= 4 && " + "document.querySelector('#status-text')?.textContent === 'dashboard ready'", + timeout=180_000, + ) + assert page.locator("#tbl-dash-ladder tbody tr").count() >= 1 + assert "Plotly could not be loaded" in page.locator("#plot-dash-ladder").inner_text() + + # Exercise a deliberate dashboard run and prove it traversed + # the Pyodide bridge, then switch panels and return. + page.evaluate( + """() => { + const original = window.flylabCall; + window.__smokeCalls = {}; + window.flylabCall = function (route, payload) { + const key = route.split('?')[0]; + window.__smokeCalls[key] = (window.__smokeCalls[key] || 0) + 1; + return original.call(this, route, payload); + }; + }""" + ) + page.locator("button[data-run='dashboard']").click() + page.wait_for_function( + "() => (window.__smokeCalls['/api/dashboard'] || 0) === 1 && " + "document.querySelector('#status-text')?.textContent === 'dashboard ready'", + timeout=180_000, + ) + + # Portable setup files round-trip the run controls and the + # experiment form without carrying stale calculated output. + page.locator("#f-compound").select_option("fipronil") + page.locator("#f-conc").fill("3e-7") + page.locator("#f-conc").dispatch_event("change") + page.locator("#f-exp-compounds").select_option(["fipronil"]) + page.locator("#f-exp-start").fill("1e-8") + page.locator("#f-exp-start").dispatch_event("change") + page.locator("#f-exp-stop").fill("1e-5") + page.locator("#f-exp-stop").dispatch_event("change") + page.locator("#f-exp-ppd").fill("2") + page.locator("#f-exp-ppd").dispatch_event("change") + page.locator("#f-exp-reps").fill("4") + page.locator("#f-exp-reps").dispatch_event("change") + page.locator("#f-exp-assay").select_option("taste_map") + page.locator("#f-exp-readouts").select_option(["mn9_hz", "mean_hz"]) + page.locator("#f-exp-randomize").check() + page.locator("#f-exp-blind").check() + + with page.expect_download() as download_info: + page.locator("#btn-preset-export").click() + download = download_info.value + preset_path = tmp_path / download.suggested_filename + download.save_as(preset_path) + preset = json.loads(preset_path.read_text()) + assert preset["format"] == "flylab.design-preset" + assert preset["version"] == 1 + assert preset["design"]["compound"] == "fipronil" + assert preset["experiment"]["assay"] == "taste_map" + assert preset["experiment"]["replicates"] == 4 + + page.locator("#f-compound").select_option("imidacloprid") + page.locator("#f-exp-assay").select_option("subgraph") + page.locator("#f-exp-reps").fill("2") + page.locator("#f-exp-reps").dispatch_event("change") + page.locator("#f-preset-file").set_input_files(str(preset_path)) + page.locator("#preset-status").wait_for(state="visible") + page.wait_for_function("() => document.querySelector('#preset-status').textContent.startsWith('Imported preset.')") + assert page.locator("#f-compound").input_value() == "fipronil" + assert float(page.locator("#f-conc").input_value()) == pytest.approx(3e-7) + assert page.locator("#f-exp-assay").input_value() == "taste_map" + assert page.locator("#f-exp-reps").input_value() == "4" + assert page.locator("#f-exp-randomize").is_checked() + assert page.locator("#f-exp-blind").is_checked() + + # Invalid files must not partially change the active design. + invalid_path = tmp_path / "invalid-preset.json" + invalid_path.write_text(json.dumps({"format": "wrong"})) + page.locator("#f-preset-file").set_input_files(str(invalid_path)) + page.wait_for_function("() => document.querySelector('#preset-status').textContent.startsWith('Import failed:')") + assert page.locator("#f-compound").input_value() == "fipronil" + + page.locator("#tab-notebook").click() + assert page.locator("#tab-notebook").get_attribute("aria-selected") == "true" + assert "active" in page.locator("#panel-notebook").get_attribute("class") + page.locator("#tab-dashboard").click() + assert page.locator("#tab-dashboard").get_attribute("aria-selected") == "true" + assert "active" in page.locator("#panel-dashboard").get_attribute("class") + + # At phone width, the grouped native picker replaces the two + # horizontal tab rows and stays synced with the active panel. + page.set_viewport_size({"width": 390, "height": 844}) + assert page.locator("#mobile-panel-picker").is_visible() + assert page.locator("#main > .tabs").first.is_hidden() + page.locator("#mobile-panel-picker").select_option("controls") + assert "active" in page.locator("#panel-controls").get_attribute("class") + assert page.locator("#tab-controls").get_attribute("aria-selected") == "true" + finally: + browser.close() + finally: + server.shutdown() + server.server_close() + server_thread.join(timeout=5) diff --git a/tests/test_build_pages.py b/tests/test_build_pages.py index 5f465bb..c2aedec 100644 --- a/tests/test_build_pages.py +++ b/tests/test_build_pages.py @@ -24,6 +24,21 @@ spec.loader.exec_module(build_pages) +@pytest.fixture(autouse=True) +def local_chart_vendor(monkeypatch: pytest.MonkeyPatch): + """Keep layout tests offline while still checking vendored asset paths.""" + def write_stub_assets(out: Path) -> list[str]: + dest = out / "vendor" + dest.mkdir(parents=True, exist_ok=True) + paths = [] + for name in build_pages.JS_VENDOR: + (dest / name).write_text("/* test-only chart vendor stub */\n") + paths.append(f"vendor/{name}") + return paths + + monkeypatch.setattr(build_pages, "vendor_js", write_stub_assets) + + @pytest.fixture def stub_wheel(tmp_path: Path) -> Path: """A wheel-shaped zip: enough for the layout checks, free to make.""" @@ -46,6 +61,8 @@ def test_site_has_every_file_the_bench_needs(site): out, _ = site for name in ("index.html", "app.js", "styles.css", "flylab-boot.js", "manifest.json"): assert (out / name).is_file(), name + for name in build_pages.JS_VENDOR: + assert (out / "vendor" / name).is_file(), name assert (out / "wheels" / "flylab-0.5.0-py3-none-any.whl").is_file() for rel in build_pages.DATA_FILES: assert (out / rel).is_file(), rel diff --git a/tests/test_pharm_assay_gains.py b/tests/test_pharm_assay_gains.py index 5794556..c1fb297 100644 --- a/tests/test_pharm_assay_gains.py +++ b/tests/test_pharm_assay_gains.py @@ -52,14 +52,13 @@ def test_taste_still_warns_about_gains_it_really_cannot_inject(): assert any("g_nav" in w and "not applied" in w for w in nb["warnings"]) -def test_ache_inhibitor_reaches_the_taste_circuit(): - """Low AChE inhibition raises ACh tone and MN9 drive; high inhibition blocks it.""" +def test_unsourced_ache_parameter_cannot_change_the_taste_circuit(): + """Withheld chlorpyrifos-oxon potency must yield no insect AChE gain.""" low = run_taste_assay("chlorpyrifos_oxon", 3e-9, sugar_hz=20.0) high = run_taste_assay("chlorpyrifos_oxon", 1e-6, sugar_hz=20.0) - assert low["gains"]["ach_tone"] > 1.0 - assert high["gains"]["ach_tone"] > low["gains"]["ach_tone"] - assert low["readouts"]["mn9_sugar_hz"] > low["readouts"]["mn9_vehicle_sugar_hz"] - assert high["readouts"]["mn9_sugar_hz"] < high["readouts"]["mn9_vehicle_sugar_hz"] + assert low["gains"]["ach_tone"] == high["gains"]["ach_tone"] == 1.0 + assert low["readouts"]["mn9_sugar_hz"] == low["readouts"]["mn9_vehicle_sugar_hz"] + assert high["readouts"]["mn9_sugar_hz"] == high["readouts"]["mn9_vehicle_sugar_hz"] @census diff --git a/tests/test_pharm_evidence.py b/tests/test_pharm_evidence.py index 696fee6..df072e6 100644 --- a/tests/test_pharm_evidence.py +++ b/tests/test_pharm_evidence.py @@ -208,7 +208,8 @@ def test_describe_is_a_provenance_record(): assert record["relation"].startswith("exact_compound_exact_receptor") assert record["species"] assert record["evidence_tier"] == "literature_order" - assert "Ratra" in record["source"] + assert "Hainzl" in record["source"] + assert "multi-species binding-site proxy" in record["source"] assert record["modelled"] is True # the aphid Kd: a genuine binding constant, but not measured in the fly diff --git a/tests/test_pharm_expression.py b/tests/test_pharm_expression.py index a4e4445..45e4e6b 100644 --- a/tests/test_pharm_expression.py +++ b/tests/test_pharm_expression.py @@ -27,13 +27,24 @@ def test_expression_table_rows_and_warning(): assert row["mapping"] == "cross_atlas_inference" assert EXPRESSION_WARNING in table["warnings"] assert table["motor_neuron_gap"]["status"] == "not_sourced" + public_lead = table["motor_neuron_gap"]["public_data_lead"] + assert public_lead["accession"] == "GSE141807" + assert public_lead["status"] == "reanalyzed_vnc_wide_cell_classes_unresolved" + assert public_lead["summary_file"] == "vnc_receptor_detection_GSE141807.yaml" + assert "not the brain subesophageal zone" in public_lead["scope"] def test_coverage_is_reported_and_incomplete(): cov = expression_table()["coverage"] - assert 0.0 < cov["fraction_known_overall"] < 1.0 + pair_fraction = cov["fraction_cell_receptor_pairs_annotated"] + assert 0.0 < pair_fraction < 1.0 + assert cov["coverage_unit"] == "graph_node_x_receptor_key_pair" + assert cov["fraction_known_overall"] == pair_fraction # legacy alias + assert cov["n_annotated_cell_receptor_pairs"] / cov["n_cell_receptor_pairs"] == pytest.approx(pair_fraction) + assert 0.0 < cov["fraction_graph_nodes_with_any_class_annotation"] < 1.0 named = cov["graphs"]["named"] assert named["n_cells"] > 1000 + assert 0 <= named["n_cells_with_any_class_annotation"] <= named["n_cells"] for receptor in ("insect_nAChR", "insect_RDL", "insect_GluCl"): block = named[receptor] assert block["n_known"] + block["n_unknown"] == named["n_cells"] diff --git a/tests/test_pharm_library.py b/tests/test_pharm_library.py index ba052f6..8a8b04d 100644 --- a/tests/test_pharm_library.py +++ b/tests/test_pharm_library.py @@ -233,15 +233,15 @@ def test_the_native_drosophila_key_is_documented_as_a_preparation(): assert "insect_nAChR_native_dmel" not in receptors["insect_nAChR"]["subunit_resolved_keys"] -def test_subunit_split_puts_each_compound_where_its_source_measured(): +def test_subunit_split_keeps_only_supported_spinosad_target_metadata(): rows = {k: set(v["receptors"]) for k, v in LIB["compounds"].items()} - # spinosad's number comes from Dalpha6 work, so it sits on the alpha6 key + # Dalpha6 target genetics do not establish a 5 uM receptor EC50. assert "insect_nAChR_alpha6" in rows["spinosad"] assert "insect_nAChR_beta1" not in rows["spinosad"] - # ... and the aggregate row keeps the same value, so the gain rules still fire alpha6 = LIB["compounds"]["spinosad"]["receptors"]["insect_nAChR_alpha6"] aggregate = LIB["compounds"]["spinosad"]["receptors"]["insect_nAChR"] - assert alpha6["value_M"] == aggregate["value_M"] == 5.0e-06 + assert alpha6["value_M"] is aggregate["value_M"] is None + assert alpha6["relation"] == aggregate["relation"] == "unsupported" assert "alpha6" in aggregate["source"].lower() or "dalpha6" in aggregate["source"].lower() # no neonicotinoid or sulfoximine claims an alpha6 row for key, entry in LIB["compounds"].items(): diff --git a/tests/test_pharm_mechanisms.py b/tests/test_pharm_mechanisms.py index 71c30f3..f6dfa3f 100644 --- a/tests/test_pharm_mechanisms.py +++ b/tests/test_pharm_mechanisms.py @@ -103,10 +103,13 @@ def test_library_compounds_reach_each_gain(): "g_glu": gains_from_occupancy(compare_compound("ivermectin", 1e-6)["receptors"])["g_glu"], "g_nav": gains_from_occupancy(compare_compound("deltamethrin", 1e-6)["receptors"])["g_nav"], "g_oct": gains_from_occupancy(compare_compound("chlordimeform", 1e-4)["receptors"])["g_oct"], - "ach_tone": gains_from_occupancy(compare_compound("chlorpyrifos_oxon", 1e-7)["receptors"])["ach_tone"], } for key, value in reached.items(): assert value != 1.0, f"{key} never leaves 1.0" + # The cited review does not establish a Drosophila oxon IC50, so this + # compound must not activate the AChE-derived gain until it is sourced. + oxon = gains_from_occupancy(compare_compound("chlorpyrifos_oxon", 1e-7)["receptors"]) + assert oxon["ach_tone"] == 1.0 def test_mechanism_table_is_renderable(): diff --git a/tests/test_server.py b/tests/test_server.py index d7f49e1..79ad5bc 100644 --- a/tests/test_server.py +++ b/tests/test_server.py @@ -728,7 +728,7 @@ def test_trust_panel_is_per_layer_and_never_blended(imidacloprid_dashboard): } <= areas assert trust["blended_confidence"] is None expression = next(r for r in trust["rows"] if r["area"] == "Receptor expression") - assert "20." in expression["status"] # 20.5% mapped + assert "20." in expression["status"] # graph-node × receptor-key annotation availability topology = next(r for r in trust["rows"] if r["area"] == "Circuit topology") assert topology["classification"] == "LITERATURE" transmitters = next(r for r in trust["rows"] if r["area"] == "Transmitter identity") diff --git a/tests/test_validation_rank.py b/tests/test_validation_rank.py index 1c853c7..4c84257 100644 --- a/tests/test_validation_rank.py +++ b/tests/test_validation_rank.py @@ -140,17 +140,17 @@ def test_report_markdown_has_a_table_and_the_discrepancies(): # the v0.5 library correction # -------------------------------------------------------------------------- def test_fipronil_vertebrate_correction_is_in_compare_compound(): - """Ratra & Casida 2001 report 1103 nM, not the 1.0e-5 M v0.4 carried.""" + """The 1103 nM entry is a multi-species radioligand result.""" rows = {r["receptor"]: r for r in compare_compound("fipronil", 1e-6)["receptors"]} vert = rows["vertebrate_GABA_A"] assert vert["ec50_M"] == pytest.approx(1.1e-6) assert "1103" in vert["source"] - assert "beta3" in vert["source"] # the homopentamer caveat is stated + assert "Hainzl" in vert["source"] and "multi-species" in vert["source"] assert vert["evidence_tier"] == "literature_order" - # the insect row keeps its value and now records the 1500-fold spread + # The insect number is explicitly labelled an assay-transfer proxy. insect = rows["insect_RDL"] assert insect["ec50_M"] == pytest.approx(3.0e-8) - assert "Lees" in insect["source"] and "house-fly" in insect["source"] + assert "Lees" in insect["source"] and "assay-transfer proxy" in insect["source"] def test_fipronil_selectivity_direction_and_gains_survive_the_correction(): diff --git a/tests/test_vnc_expression_reanalysis.py b/tests/test_vnc_expression_reanalysis.py new file mode 100644 index 0000000..9a83846 --- /dev/null +++ b/tests/test_vnc_expression_reanalysis.py @@ -0,0 +1,29 @@ +from __future__ import annotations + +from io import BytesIO + +from scripts.reanalyze_vnc_expression import analyze_matrix + + +def test_vnc_matrix_qc_and_gene_detection_use_published_cutoffs(): + barcodes = ["pass", "too_few_genes", "at_umi_floor", "high_mito"] + rows = ["GENE\t" + "\t".join(barcodes)] + for index in range(200): + rows.append(f"gene{index:03d}\t1\t1\t1\t1") + # The nicotinic row brings barcode 1 above 1,200 UMIs. Barcode 2 has only + # 200 genes; barcode 3 sits exactly at the excluded 1,200-UMI floor. + rows.append("nAChRalpha1\t1100\t0\t1000\t1001") + rows.append("mt:ND1\t0\t0\t0\t300") + matrix = BytesIO(("\n".join(rows) + "\n").encode()) + + result = analyze_matrix(matrix, "GSM4213596_female_rep1_dge.txt.gz") + + assert result["cells_in_matrix"] == 4 + assert result["cells_after_published_qc"] == 1 + assert result["published_cells_after_qc"] == 2590 + assert result["cell_count_difference"] == -2589 + assert result["targets"]["insect_nAChR"]["genes"]["nAChRalpha1"] == { + "n_detected": 1, + "fraction_detected": 1.0, + } + assert result["targets"]["insect_RDL"]["n_genes_matched"] == 0