Machine Learning Coarse-Grained Potentials of Protein Thermodynamics

TL;DR

This paper introduces machine learning-based coarse-grained potentials derived from neural networks trained on extensive all-atom simulations, enabling faster and thermodynamically accurate protein dynamics modeling.

Contribution

It presents a novel neural network approach to create coarse-grained potentials that accelerate simulations while maintaining thermodynamic fidelity across multiple proteins.

Findings

Accelerates protein dynamics by over 1000 times

Preserves thermodynamic properties of all-atom systems

Captures structural features of mutated proteins

Abstract

A generalized understanding of protein dynamics is an unsolved scientific problem, the solution of which is critical to the interpretation of the structure-function relationships that govern essential biological processes. Here, we approach this problem by constructing coarse-grained molecular potentials based on artificial neural networks and grounded in statistical mechanics. For training, we build a unique dataset of unbiased all-atom molecular dynamics simulations of approximately 9 ms for twelve different proteins with multiple secondary structure arrangements. The coarse-grained models are capable of accelerating the dynamics by more than three orders of magnitude while preserving the thermodynamics of the systems. Coarse-grained simulations identify relevant structural states in the ensemble with comparable energetics to the all-atom systems. Furthermore, we show that a single coarse-grained potential can integrate all twelve proteins and can capture experimental structural features of mutated proteins. These results indicate that machine learning coarse-grained potentials could provide a feasible approach to simulate and understand protein dynamics.