Learning Pair Potentials using Differentiable Simulations

TL;DR

This paper introduces a flexible, gradient-based method called DiffSim for learning pair interaction potentials directly from structural data using differentiable molecular dynamics simulations, enabling simultaneous multi-condition fitting.

Contribution

The paper presents a novel stochastic approach using differentiable simulations to learn pair potentials from data, allowing multi-system optimization and improved potential transferability.

Findings

Successfully recovered Lennard-Jones potentials from radial distribution functions.

DiffSim outperforms traditional methods like Iterative Boltzmann Inversion in exploring potential space.

Able to fit potentials across different temperatures and compositions simultaneously.

Abstract

Learning pair interactions from experimental or simulation data is of great interest for molecular simulations. We propose a general stochastic method for learning pair interactions from data using differentiable simulations (DiffSim). DiffSim defines a loss function based on structural observables, such as the radial distribution function, through molecular dynamics (MD) simulations. The interaction potentials are then learned directly by stochastic gradient descent, using backpropagation to calculate the gradient of the structural loss metric with respect to the interaction potential through the MD simulation. This gradient-based method is flexible and can be configured to simulate and optimize multiple systems simultaneously. For example, it is possible to simultaneously learn potentials for different temperatures or for different compositions. We demonstrate the approach by recovering simple pair potentials, such as Lennard-Jones systems, from radial distribution functions. We find that DiffSim can be used to probe a wider functional space of pair potentials compared to traditional methods like Iterative Boltzmann Inversion. We show that our methods can be used to simultaneously fit potentials for simulations at different compositions and temperatures to improve the transferability of the learned potentials.