BioinfoMachineLearning/PoseBench

Comprehensive benchmarking of protein-ligand structure prediction methods. (Nature Machine Intelligence)

65
/ 100
Established

Provides standardized evaluation across 10+ structure prediction methods (DiffDock, FABind, RoseTTAFold-All-Atom, Chai-1, Boltz, and others) using Hydra-based configuration management for reproducible experimental sweeps and ensemble inference. Includes curated benchmark datasets (Astex Diverse, DockGen-E, PoseBusters) with automated scoring pipelines and comparative visualization tools, available as both a pip-installable package and full development environment with pre-configured method implementations.

213 stars and 148 monthly downloads. Available on PyPI.

Maintenance 13 / 25
Adoption 15 / 25
Maturity 25 / 25
Community 12 / 25

How are scores calculated?

Stars

213

Forks

16

Language

Jupyter Notebook

License

MIT

Last pushed

Mar 13, 2026

Monthly downloads

148

Commits (30d)

0

Dependencies

38

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/BioinfoMachineLearning/PoseBench"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.