Physics-aware Graph Neural Network for Accurate RNA 3D Structure Prediction

TL;DR

This paper introduces PaxNet, a physics-aware graph neural network that improves RNA 3D structure prediction accuracy by modeling local and non-local interactions, outperforming existing methods on benchmark datasets.

Contribution

The work presents a novel GNN architecture that incorporates physics-inspired interaction modeling and attention-based fusion for RNA structure prediction, trained with limited data.

Findings

PaxNet significantly outperforms state-of-the-art baselines.

The model effectively captures local and non-local interactions.

Results demonstrate potential for broader application to macromolecular structure modeling.

Abstract

Biological functions of RNAs are determined by their three-dimensional (3D) structures. Thus, given the limited number of experimentally determined RNA structures, the prediction of RNA structures will facilitate elucidating RNA functions and RNA-targeted drug discovery, but remains a challenging task. In this work, we propose a Graph Neural Network (GNN)-based scoring function trained only with the atomic types and coordinates on limited solved RNA 3D structures for distinguishing accurate structural models. The proposed Physics-aware Multiplex Graph Neural Network (PaxNet) separately models the local and non-local interactions inspired by molecular mechanics. Furthermore, PaxNet contains an attention-based fusion module that learns the individual contribution of each interaction type for the final prediction. We rigorously evaluate the performance of PaxNet on two benchmarks and compare it with several state-of-the-art baselines. The results show that PaxNet significantly outperforms all the baselines overall, and demonstrate the potential of PaxNet for improving the 3D structure modeling of RNA and other macromolecules. Our code is available at https://github.com/zetayue/Physics-aware-Multiplex-GNN.