Learning Large-Time-Step Molecular Dynamics with Graph Neural Networks

TL;DR

This paper introduces MDNet, a graph neural network model that predicts atomic trajectories with large time steps, significantly improving efficiency while maintaining accuracy in molecular dynamics simulations.

Contribution

The paper presents MDNet, a GNN-based model capable of large-time-step molecular dynamics prediction, scalable to large systems, and aligned with standard MD results.

Findings

MDNet accurately predicts equilibrium properties.

MDNet maintains transport properties.

MDNet scales linearly with system size.

Abstract

Molecular dynamics (MD) simulation predicts the trajectory of atoms by solving Newton's equation of motion with a numeric integrator. Due to physical constraints, the time step of the integrator need to be small to maintain sufficient precision. This limits the efficiency of simulation. To this end, we introduce a graph neural network (GNN) based model, MDNet, to predict the evolution of coordinates and momentum with large time steps. In addition, MDNet can easily scale to a larger system, due to its linear complexity with respect to the system size. We demonstrate the performance of MDNet on a 4000-atom system with large time steps, and show that MDNet can predict good equilibrium and transport properties, well aligned with standard MD simulations.