Exploring Optimal Transport-Based Multi-Grained Alignments for Text-Molecule Retrieval

TL;DR

This paper introduces ORMA, a novel multi-grained alignment model using optimal transport and contrastive learning to improve text-molecule retrieval by capturing hierarchical sub-structure details.

Contribution

The paper presents the first multi-grained alignment approach at motif and multi-token levels for text-molecule retrieval, leveraging optimal transport and hierarchical molecule modeling.

Findings

ORMA outperforms state-of-the-art models on ChEBI-20 and PCdes datasets.

Multi-grained alignments improve retrieval accuracy.

Optimal transport effectively aligns tokens with molecular motifs.

Abstract

The field of bioinformatics has seen significant progress, making the cross-modal text-molecule retrieval task increasingly vital. This task focuses on accurately retrieving molecule structures based on textual descriptions, by effectively aligning textual descriptions and molecules to assist researchers in identifying suitable molecular candidates. However, many existing approaches overlook the details inherent in molecule sub-structures. In this work, we introduce the Optimal TRansport-based Multi-grained Alignments model (ORMA), a novel approach that facilitates multi-grained alignments between textual descriptions and molecules. Our model features a text encoder and a molecule encoder. The text encoder processes textual descriptions to generate both token-level and sentence-level representations, while molecules are modeled as hierarchical heterogeneous graphs, encompassing atom, motif, and molecule nodes to extract representations at these three levels. A key innovation in ORMA is the application of Optimal Transport (OT) to align tokens with motifs, creating multi-token representations that integrate multiple token alignments with their corresponding motifs. Additionally, we employ contrastive learning to refine cross-modal alignments at three distinct scales: token-atom, multitoken-motif, and sentence-molecule, ensuring that the similarities between correctly matched text-molecule pairs are maximized while those of unmatched pairs are minimized. To our knowledge, this is the first attempt to explore alignments at both the motif and multi-token levels. Experimental results on the ChEBI-20 and PCdes datasets demonstrate that ORMA significantly outperforms existing state-of-the-art (SOTA) models.