PoseBusters: Plausibility checks for generated molecule poses.
Paper in Chemical Science and preprint on arXiv
# install with pip from PyPI
pip install posebusters
# Check generated molecule pose.
bust molecule_pred.sdf
bust molecule_a.sdf molecule_b.sdf
bust molecule_*.sdf
# Check new ligand generated for a given protein.
bust ligand_pred.sdf -p mol_cond.pdb
# Check re-docked ligand (a pose that should recover the ligand in a given protein-ligand crystal complex).
bust ligand_pred.sdf -l mol_true.sdf -p protein.pdb
# Check any of the three by providing a csv with files to check together
bust -t file_table.csv
Documentation is available at https://posebusters.readthedocs.io.
For more information about the tests and for a study using PoseBusters to compare docking methods, refer to our paper or preprint:
@article{buttenschoen2023posebusters,
title = {{{PoseBusters}}: {{AI-based}} Docking Methods Fail to Generate Physically Valid Poses or Generalise to Novel Sequences},
shorttitle = {{{PoseBusters}}},
author = {Buttenschoen, Martin and Morris, Garrett M. and Deane, Charlotte M.},
year = "2023",
publisher = "The Royal Society of Chemistry",
doi = "10.1039/D3SC04185A",
url = "https://dx.doi.org/10.1039/D3SC04185A",
}
The data used for the paper is available at https://zenodo.org/record/8278563.
We welcome all feedback. For code issues, please open an issue. For other inquiries contact us by email.
This program uses software written by other people. Notably: