Code to generate the TraitBlender-DCM-Benchmarks dataset: phylogenetic trait simulations (Brownian motion, Pagel’s λ, regime OU), TraitBlender / MorphoWeave imaging configs, and Slurm drivers. The released trees, tip tables, meshes, and renders are on Hugging Face.
Dataset | TraitBlender | Docs | Zenodo
- Overview
- Dataset
- Repository layout
- Installation
- Running simulations
- Running imaging
- Citation
- License
- Acknowledgements
This pipeline simulates tip PCs under known comparative processes on a shared ultrametric balanced tree ((T = 1)), then renders three MorphoWeave morphospaces (mouse cranium, primate molar, squamate vertebra) with TraitBlender. The Hugging Face release is 3 scenarios × 3 morphospaces × 10 parameter levels × 1 replicate (rep_01) = 90 imaged simulations (128 tips × 3 views × 5 frames = 1,920 PNGs per simulation).
Released data: imageomics/TraitBlender-DCM-Benchmarks (dataset card on the Hub).
To regenerate or extend the release, set DATA_ROOT to a local data directory and SSM_ROOT to your MorphoWeave SSM libraries (see below). MorphoWeave SSMs are not redistributed with the Hub dataset.
Related software: TraitBlender (docs, Zenodo).
.
README.md
CITATION.cff
LICENSE.md
environment.yml
configs/
tabular/ # balanced_bm / balanced_ou / pagel_lambda_bm
traitblender/ # one MorphoWeave template YAML per morphospace
scripts/
sim_traits/ # simulate_traits.R + helpers.R
imaging/ # make_jobs.sh + Blender Python driver
slurm/ # array job templates
manifests/
tabular.txt
imaging.txt
Job ids: {tabular_id}__{template_id}__{sim} (e.g. balanced_bm__vertebra_3000_smooth__sig2_01__rep_01).
To add a morphospace, add configs/traitblender/<id>.yaml and include <id> in TEMPLATES when running make_jobs.sh.
conda env create -f environment.yml
conda activate tb-dcm-benchmarks
export CODE_ROOT=/path/to/TraitBlender-DCM-Benchmarks
export DATA_ROOT=/path/to/data
export SSM_ROOT=/path/to/morphoweave_ssmsSoftware dependencies are pinned in environment.yml (R + ape, phytools, geiger, yaml). Imaging also requires Blender 5.1+ with TraitBlender installed.
Each tabular YAML sets seed: for reproducibility. Pass Slurm account / partition / qos on the sbatch command line—do not hard-code them in the checked-in scripts. Edit scripts/slurm/run_imaging.slurm module loads for your cluster.
cd "$CODE_ROOT"
Rscript scripts/sim_traits/simulate_traits.R \
configs/tabular/balanced_bm.yaml "$DATA_ROOT"
mkdir -p "${CODE_ROOT}/logs"
sbatch \
--account=YOUR_ACCOUNT \
--partition=YOUR_PARTITION \
--qos=YOUR_QOS \
--output="${CODE_ROOT}/logs/%x_%A_%a.out" \
--error="${CODE_ROOT}/logs/%x_%A_%a.err" \
"${CODE_ROOT}/scripts/slurm/run_tabular_sims.slurm"Configs use 10 parameter levels with n_rep: 1 for the HF release. Outputs under $DATA_ROOT:
trees/balanced_128.nwk
balanced_bm/
tabular/sig2_XX__rep_YY.csv
tabular_tips/
trees/sig2_XX__rep_YY.nwk
params.csv
run_meta.yaml
renders/<morphospace>/<sim>/…
balanced_ou/
tabular/alpha_XX__rep_YY.csv
trees/alpha_XX__rep_YY.nex
…
pagel_lambda_bm/
tabular/lambda_XX__rep_YY.csv
trees/lambda_XX.nwk
…
bash "$CODE_ROOT/scripts/imaging/make_jobs.sh"
N=$(grep -cve '^\s*(#|$)' "$CODE_ROOT/manifests/imaging.txt")
sbatch --account=... --partition=... --qos=... --array=1-"$N" \
--output="$CODE_ROOT/logs/%x_%A_%a.out" \
--error="$CODE_ROOT/logs/%x_%A_%a.err" \
"$CODE_ROOT/scripts/slurm/run_imaging.slurm"make_jobs.sh writes tip-only CSVs under $DATA_ROOT and filled TraitBlender job YAMLs under configs/traitblender/jobs/ (gitignored).
If you use this software, please cite it (see also the “Cite this repository” button from CITATION.cff). Please also cite the dataset (and the forthcoming dataset paper when available), and TraitBlender when using the imaging stack.
Data
@misc{charpentier_traitblender_dcm_benchmarks,
author = {Charpentier, Caleb P.},
title = {{TraitBlender-{DCM}-Benchmarks}},
year = {2026},
url = {https://huggingface.co/datasets/imageomics/TraitBlender-DCM-Benchmarks},
publisher = {Hugging Face}
}Code (this repository)
@software{charpentier_traitblender_dcm_benchmarks_code,
author = {Charpentier, Caleb P.},
title = {{TraitBlender-{DCM}-Benchmarks} (code)},
year = {2026},
url = {https://github.com/Imageomics/TraitBlender-DCM-Benchmarks}
}@software{charpentier_2026_traitblender,
author = {Charpentier, Caleb and Linscott, T. Mason and Bradley, John and Campolongo, Elizabeth and Thompson, Matthew and Porto, Arthur and Uyeda, Josef C.},
title = {TraitBlender},
year = {2026},
publisher = {Zenodo},
doi = {10.5281/zenodo.14804133},
url = {https://doi.org/10.5281/zenodo.14804133}
}Code is MIT. The Hugging Face dataset is CC BY 4.0. TraitBlender / MorphoWeave / Blender retain their own licenses.
This work was supported by the Imageomics Institute, which is funded by the US National Science Foundation's Harnessing the Data Revolution (HDR) program under Award #2118240 (Imageomics: A New Frontier of Biological Information Powered by Knowledge-Guided Machine Learning). Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.