Repository for the ANA2B publication.
Weights for the models necessary to reproduce results can be found in source/weights_models/
Necessary datasets (need to be downloaded separately):
For Training:
- DES5M: https://www.nature.com/articles/s41597-021-00833-x
- AlphaML: https://www.nature.com/articles/s41597-019-0157-8
- Intramolecular Gradients/Multipoles (PBE0/def2-TZVP): https://www.research-collection.ethz.ch/handle/20.500.11850/626683
For Validation and Testing:
- BioFragmentDB: http://vergil.chemistry.gatech.edu/active_bfdb/bfdb/cgi-bin/bfdb.py
- S7L: https://www.nature.com/articles/s41467-021-24119-3
- ICE13: https://pubs.aip.org/aip/jcp/article/157/13/134701/2841942/
- X23: https://pubs.rsc.org/en/content/articlelanding/2019/cp/c9cp04488d
- Structures 6th Blindtest: https://www.science.org/doi/10.1126/sciadv.aau3338
- Structures 5th Blindtest: https://journals.iucr.org/b/issues/2011/06/00/bk5106/index.html
- Structures 4th Blindtest: https://journals.iucr.org/b/issues/2009/02/00/bk5081/index.html
- Structures 3rd Blindtest: https://journals.iucr.org/b/issues/2005/05/00/de5014/index.html
- Structures 2nd Blindtest: https://onlinelibrary.wiley.com/iucr/doi/10.1107/S0108768102005669
- Structures 1st Blindtest: https://onlinelibrary.wiley.com/iucr/doi/10.1107/S0108768100004584
Examples and scripts necessary to reproduce results can be found in
- MD: source/anaff/md
- CSP: source/anaff/crystal_ranking
- Dimers: source/results
- ANA2B model training: source/anaff/train_model_alpha_gnn.py
- A small example for general usage: source/examples/example.ipynb
After cloning the repo, following package are required:
- Tensorflow
- ase
- mdtraj
- graph_nets
- Simple DFT-D3
Exemplary usecases can be found in
- source/examples/example.ipynb
- source/anaff/md/run_md_ase.py
@article{ANA2B
author ="Thürlemann, Moritz and Riniker, Sereina",
title ="Hybrid classical/machine-learning force fields for the accurate description of molecular condensed-phase systems",
journal ="Chem. Sci.",
year ="2023",
doi ="10.1039/D3SC04317G",
}
MIT