Ewald-based Long-Range Message Passing for Molecular Graphs

TL;DR

This paper introduces Ewald message passing, a Fourier space method to incorporate long-range interactions into molecular graph neural networks, improving energy predictions especially for structures with significant long-range effects.

Contribution

It proposes a novel Ewald-based message passing scheme that enhances existing MPNNs by efficiently modeling long-range interactions in molecular data.

Findings

Consistently improves energy MAE across models and datasets.

Achieves around 10-16% reduction in error.

Significantly benefits structures with high long-range energy contributions.

Abstract

Neural architectures that learn potential energy surfaces from molecular data have undergone fast improvement in recent years. A key driver of this success is the Message Passing Neural Network (MPNN) paradigm. Its favorable scaling with system size partly relies upon a spatial distance limit on messages. While this focus on locality is a useful inductive bias, it also impedes the learning of long-range interactions such as electrostatics and van der Waals forces. To address this drawback, we propose Ewald message passing: a nonlocal Fourier space scheme which limits interactions via a cutoff on frequency instead of distance, and is theoretically well-founded in the Ewald summation method. It can serve as an augmentation on top of existing MPNN architectures as it is computationally inexpensive and agnostic to architectural details. We test the approach with four baseline models and two datasets containing diverse periodic (OC20) and aperiodic structures (OE62). We observe robust improvements in energy mean absolute errors across all models and datasets, averaging 10% on OC20 and 16% on OE62. Our analysis shows an outsize impact of these improvements on structures with high long-range contributions to the ground truth energy.