Extracting Molecular Properties from Natural Language with Multimodal Contrastive Learning

TL;DR

This paper explores how to transfer molecular property knowledge from natural language descriptions to graph representations using contrastive learning, improving property prediction accuracy in computational biochemistry.

Contribution

It introduces a novel contrastive learning approach aligning text and graph representations, along with new graph augmentation strategies, to enhance molecular property prediction.

Findings

Achieved +4.26% AUROC improvement over graph-only models

Gained +1.54% AUROC over existing contrastive models

Enhanced text retrieval with neural relevance scoring

Abstract

Deep learning in computational biochemistry has traditionally focused on molecular graphs neural representations; however, recent advances in language models highlight how much scientific knowledge is encoded in text. To bridge these two modalities, we investigate how molecular property information can be transferred from natural language to graph representations. We study property prediction performance gains after using contrastive learning to align neural graph representations with representations of textual descriptions of their characteristics. We implement neural relevance scoring strategies to improve text retrieval, introduce a novel chemically-valid molecular graph augmentation strategy inspired by organic reactions, and demonstrate improved performance on downstream MoleculeNet property classification tasks. We achieve a +4.26% AUROC gain versus models pre-trained on the graph modality alone, and a +1.54% gain compared to recently proposed molecular graph/text contrastively trained MoMu model (Su et al. 2022).