biased-force-matching

This repository contains supporting code accompanying the paper:

Enhanced Sampling for Efficient Training of Coarse Grained Machine Learning Potentials Journal of Chemical Theory and Computation (JCTC), 2026
DOI: 10.1021/acs.jctc.5c01712

Figure: Table of contents graphic from the paper.

Installation

The core setup and source code are located in the chemtrain repository.

This project depends on chemtrain with a CUDA-enabled JAX backend.
Installation is managed primarily through Conda.

First, create the environment from the provided environment.yml:

conda env create -f environment.yml
conda activate biased_fm

After activating the environment, install chemutils manually under chemutils folder:

pip install -e chemutils

If you use biased-force-matching / enhanced-sampling-force-matching, please cite the following works:

@article{chen2026enhanced,
  title={Enhanced Sampling for Efficient Learning of Coarse-Grained Machine Learning Potentials},
  author={Chen, Weilong and Görlich, Franz and Fuchs, Paul and Zavadlav, Julija},
  journal={Journal of Chemical Theory and Computation},
  volume = {22},
  number = {1},
  pages = {219-230},
  year = {2026},
  doi = {10.1021/acs.jctc.5c01712},
  URL = {https://doi.org/10.1021/acs.jctc.5c01712},
  publisher={ACS Publications}
}