Training a force field for proteins and small molecules from scratch

TL;DR

This paper introduces Garnet, a graph neural network-based force field trained on diverse quantum and experimental data, enabling automated, transferable, and accurate molecular simulations for proteins and small molecules.

Contribution

It presents a novel machine learning approach to develop force fields from scratch without relying on existing parameters, improving transferability and systematic exploration.

Findings

Garnet achieves performance comparable to existing force fields on various molecules.

It provides accurate predictions for protein-ligand binding free energies.

The double exponential potential is a flexible alternative to Lennard-Jones.

Abstract

Force fields for molecular dynamics are usually developed manually, limiting their transferability and making systematic exploration of functional forms challenging. We developed a graph neural network that assigns all force field parameters for diverse molecules using continuous atom typing. The freely-available model, called Garnet, was trained on quantum mechanical, condensed phase and protein nuclear magnetic resonance data without the use of existing parameters. The resulting force field shows comparable performance to current force fields on small molecules, folded proteins, protein complexes and disordered proteins. It shows similar results to popular approaches for relative binding free energy predictions across a range of targets. Assessing different functional forms shows that the double exponential potential is a flexible and accurate alternative to the Lennard-Jones potential. Garnet provides a platform for automated, reproducible force field discovery that brings the benefits of machine learning to classical force fields.