Image-Like Graph Representations for Improved Molecular Property Prediction
This addresses the problem of molecular property prediction for researchers in chemistry and drug discovery, offering a novel approach that could improve scalability.
The paper tackles molecular property prediction by proposing CubeMol, a fixed-dimensional stochastic representation that bypasses Graph Neural Networks (GNNs), and shows it exceeds state-of-the-art GNN performance when paired with a transformer model.
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.