Direct Molecular Polarizability Prediction with SO(3) Equivariant Local Frame GNNs

TL;DR

This paper presents an $SO(3)$-equivariant GNN that predicts molecular tensorial properties directly, leveraging local coordinate frames to improve accuracy over scalar-based methods, advancing geometry-aware molecular modeling.

Contribution

Introduces a novel $SO(3)$-equivariant GNN architecture that directly predicts tensorial molecular properties using local frames, enhancing geometric information capture.

Findings

Tensorial message passing outperforms scalar message passing.

Model accurately predicts polarizabilities in QM7-X dataset.

Advances structured, geometry-aware neural models for molecules.

Abstract

We introduce a novel equivariant graph neural network (GNN) architecture designed to predict the tensorial response properties of molecules. Unlike traditional frameworks that focus on regressing scalar quantities and derive tensorial properties from their derivatives, our approach maintains $SO(3)$-equivariance through the use of local coordinate frames. Our GNN effectively captures geometric information by integrating scalar, vector, and tensor channels within a local message-passing framework. To assess the accuracy of our model, we apply it to predict the polarizabilities of molecules in the QM7-X dataset and show that tensorial message passing outperforms scalar message passing models. This work marks an advancement towards developing structured, geometry-aware neural models for molecular property prediction.