Image-Like Graph Representations for Improved Molecular Property Prediction

TL;DR

This paper introduces CubeMol, a novel fixed-dimensional stochastic molecular representation that, when combined with transformers, outperforms traditional GNNs in molecular property prediction, offering improved scalability and bypassing GNN limitations.

Contribution

The paper proposes CubeMol, a new molecular representation that eliminates the need for GNNs and enhances prediction performance with transformer models.

Findings

CubeMol outperforms state-of-the-art GNN models.

The approach offers better scalability for molecular property prediction.

Transformers combined with CubeMol achieve superior results.

Abstract

Research into deep learning models for molecular property prediction has primarily focused on the development of better Graph Neural Network (GNN) architectures. Though new GNN variants continue to improve performance, their modifications share a common theme of alleviating problems intrinsic to their fundamental graph-to-graph nature. In this work, we examine these limitations and propose a new molecular representation that bypasses the need for GNNs entirely, dubbed CubeMol. Our fixed-dimensional stochastic representation, when paired with a transformer model, exceeds the performance of state-of-the-art GNN models and provides a path for scalability.