DMMRL: Disentangled Multi-Modal Representation Learning via Variational Autoencoders for Molecular Property Prediction

TL;DR

DMMRL introduces a variational autoencoder framework that disentangles and integrates multi-modal molecular data, significantly improving interpretability and prediction accuracy in molecular property prediction tasks.

Contribution

It proposes a novel disentangled multi-modal representation learning method using variational autoencoders with regularizations and attention mechanisms for molecular data.

Findings

Outperforms state-of-the-art methods on seven benchmark datasets.

Enhances interpretability through disentangled representations.

Improves predictive accuracy by effectively integrating multi-modal information.

Abstract

Molecular property prediction constitutes a cornerstone of drug discovery and materials science, necessitating models capable of disentangling complex structure-property relationships across diverse molecular modalities. Existing approaches frequently exhibit entangled representations--conflating structural, chemical, and functional factors--thereby limiting interpretability and transferability. Furthermore, conventional methods inadequately exploit complementary information from graphs, sequences, and geometries, often relying on naive concatenation that neglects inter-modal dependencies. In this work, we propose DMMRL, which employs variational autoencoders to disentangle molecular representations into shared (structure-relevant) and private (modality-specific) latent spaces, enhancing both interpretability and predictive performance. The proposed variational disentanglement mechanism effectively isolates the most informative features for property prediction, while orthogonality and alignment regularizations promote statistical independence and cross-modal consistency. Additionally, a gated attention fusion module adaptively integrates shared representations, capturing complex inter-modal relationships. Experimental validation across seven benchmark datasets demonstrates DMMRL's superior performance relative to state-of-the-art approaches. The code and data underlying this article are freely available at https://github.com/xulong0826/DMMRL.