Skip to content

feat: make the GECKO 4 kcat and proteomics workflow the default route - #92

Open
edkerk wants to merge 43 commits into
developfrom
feat/gecko4-workflow
Open

edkerk wants to merge 43 commits into
developfrom
feat/gecko4-workflow

Conversation

@edkerk

@edkerk edkerk commented Nov 12, 2026 •

Copy link
Copy Markdown
Collaborator

The Python side of CompMetEng/GECKO#33: GECKO 3 steps that GECKO 4 replaced leave the usual workflow, and geckopy follows MATLAB where the two differed. Parity ledger: CompMetEng/raven-gecko-parity#155. Guide: CompMetEng/gecko-docs#44.

  • Functions with GECKO 3 names: geckopy started as a GECKO 4 port, so it removes them without a deprecation period. upgrade_project reports each one in a script with its replacement.
    • run_dlkcat, read_dlkcat_output, write_dlkcat_input: replaced by kcats_from_dlkcat.
    • apply_custom_kcats: replaced by data/kcats/custom.tsv.
    • get_kcat_across_isozymes, get_standard_kcat, select_kcat_value, merge_dlkcat_and_fuzzy_kcats, get_ec_from_database, get_ec_from_gem, sigma_fitter, constrain_flux_data: aliases of fill_kcats_from_isozymes, assign_standard_kcat, apply_kcat_list, merge_kcats, fill_eccodes_from_database, fill_eccodes_from_gem, fit_sigma and apply_flux_data_constraints.
    • truncate_values, find_max_value: nothing in geckopy called them, and GECKO 4 has no counterpart.
  • collect_kcats: raises ValueError on a project that has only the GECKO 3 data/customKcats.tsv, before it changes the model, and names convert_custom_kcats.
  • kcats_from_dlkcat: option via ("okp", "file", "docker"); without data/DLKcat.tsv it predicts with OpenKineticsPredictor. It returns (kcat_list, done).
  • OpenKineticsPredictor engine (submit, fetch, write input, read output): private; kcats_from_okp is the entry point.
  • okp.method, [kcat_sources.okp] methods: the default is KinForm-H alone (was CataPro, and eight methods for the presets).
  • [kcat_sources.<name>]: a table replaced the preset's settings for that source; it now overrides only the keys it sets.
  • [uniprot] id, [kegg] id: defaulted to the identifiers of Saccharomyces cerevisiae; they are now empty. A missing UniProt identifier without data/uniprot.tsv raises an error; an empty KEGG identifier skips KEGG with a warning.
  • make_ec_model: did not load UniProt and KEGG data when none was given; it now loads them with the adapter's settings and downloads a missing file, as MATLAB loadDatabases.
  • geckopy uniprot-download, download_uniprot, load_uniprot_tsv: kept only reviewed entries, against the adapter default uniprot.reviewed = false; --reviewed-only limits to reviewed entries.
  • collect_kcats: returned the model with the isozyme kcat constraints already applied, unlike MATLAB.
  • apply_complex_data: reported one complex proposal per reaction where MATLAB reports two.
  • get_complex_data: without a taxonomic id it now warns and returns [].
  • level_complex_concs: option families, as MATLAB.
  • calculate_f_factor: returned 0 without proteomics data; it now returns 0.5 with a warning, as MATLAB, and reads the project's data/paxDB.tsv when no data are given.
  • fit_sigma: accepted a target growth rate of 0 or below.
  • relax_proteomics_greedy: raised the iteration error even when the last allowed relaxation reached the target; it returns converged=False when the candidates run out.
  • get_conc_control_coeffs: continued after a first solve that was not optimal.
  • ec_fseof: returned no targets silently when production had no range to scan; it now raises ValueError. Default n_steps is 16, target labels are OE/KD/KO, and gene_targets reconciles actions and tests essentiality as MATLAB ecFSEOF. EcFSEOFResult is exported.
  • upgrade_project: names the replacement of each removed function.
  • .gitignore: ignores okpApiKey.txt.
  • Docstrings that described the code wrongly or named old MATLAB paths, and docs/relaxation_methods.md: corrected. The low-level kcat functions are marked as such.

GECKO removed truncateValues, and no function in geckopy calls
truncate_values. update_prot_pool was never ported and
merge_dlkcat_and_fuzzy_kcats is already a deprecated alias of
merge_kcats; a test now checks its warning.
kcats_from_dlkcat takes via ('okp', 'file' or 'docker') and returns
(kcat_list, done) as kcats_from_okp does. update_kcat_sources gathers the
dlkcat source through OKP when data/DLKcat.tsv is missing instead of
requiring Docker.
geckopy init writes a .gitignore with data/okpApiKey.txt in a new project and the repository ignores okpApiKey.txt. The module help of the OKP functions said the file was git-ignored, which was not the case.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant