Flow-matching -- efficient coarse-graining of molecular dynamics without forces

TL;DR

Flow-matching introduces a novel deep learning-based method for coarse-grained molecular simulations, significantly improving data efficiency and accurately capturing protein folding dynamics without relying on extensive all-atom force data.

Contribution

It presents flow-matching, a new training approach that combines normalizing flows with force matching, reducing data requirements and enhancing CG force field accuracy.

Findings

Outperforms classical force-matching by an order of magnitude in data efficiency

Successfully captures folding and unfolding transitions of small proteins

Eliminates the need for iterative all-atom simulations in CG force field training

Abstract

Coarse-grained (CG) molecular simulations have become a standard tool to study molecular processes on time- and length-scales inaccessible to all-atom simulations. Parameterizing CG force fields to match all-atom simulations has mainly relied on force-matching or relative entropy minimization, which require many samples from costly simulations with all-atom or CG resolutions, respectively. Here we present flow-matching, a new training method for CG force fields that combines the advantages of both methods by leveraging normalizing flows, a generative deep learning method. Flow-matching first trains a normalizing flow to represent the CG probability density, which is equivalent to minimizing the relative entropy without requiring iterative CG simulations. Subsequently, the flow generates samples and forces according to the learned distribution in order to train the desired CG free energy model via force matching. Even without requiring forces from the all-atom simulations, flow-matching outperforms classical force-matching by an order of magnitude in terms of data efficiency, and produces CG models that can capture the folding and unfolding transitions of small proteins.