Classical Structural Priors Improve Neural Molecular Representations in Molecule Language Models

Abstract Classical molecular similarity is central to chemical information systems, yet molecule-language models trained on paired structure–text data do not necessarily organize molecular embeddings according to graded structural relationships. Here, we introduce Structure-Aware Guided Alignment (SAGA), a training-only regularizer that uses Centered Kernel Alignment to align neural molecular representations with fingerprint-based Tanimoto similarity and requires no additional parameters or fingerprints at inference. In MolCA and Atomas, SAGA improves molecule–text retrieval, retrieves closer structural analogues, and raises average performance in scaffold-split property prediction, without clear degradation of molecular description or text-to-molecule generation. These results, together with control experiments that attribute the gains to the chemical content of the prior, show that established chemoinformatics priors can make molecule-language representations more chemically faithful while retaining their multimodal utility.

Authors

Institutions

Publication Details

Journal
Journal of Chemical Information and Modeling
Published
2026-11-16
DOI
https://doi.org/10.1021/acs.jcim.6c02225
Primary Topic
Computational Drug Discovery Methods
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Classical Structural Priors Improve Neural Molecular Representations in Molecule Language Models

Sungroh Yoon, Munyoung Chang, Donghyeok Lee, Sangkwon Park
1 links
Journal of Chemical Information and Modeling
Computational Drug Discovery Methods
article

Classical Structural Priors Improve Neural Molecular Representations in Molecule Language Models

Sungroh Yoon, Munyoung Chang, Donghyeok Lee, Sangkwon Park
article en

Abstract

Abstract Classical molecular similarity is central to chemical information systems, yet molecule-language models trained on paired structure–text data do not necessarily organize molecular embeddings according to graded structural relationships. Here, we introduce Structure-Aware Guided Alignment (SAGA), a training-only regularizer that uses Centered Kernel Alignment to align neural molecular representations with fingerprint-based Tanimoto similarity and requires no additional parameters or fingerprints at inference. In MolCA and Atomas, SAGA improves molecule–text retrieval, retrieves closer structural analogues, and raises average performance in scaffold-split property prediction, without clear degradation of molecular description or text-to-molecule generation. These results, together with control experiments that attribute the gains to the chemical content of the prior, show that established chemoinformatics priors can make molecule-language representations more chemically faithful while retaining their multimodal utility.

Journal of Chemical Information and Modeling
Seoul National University (KR), Seoul National University Bundang Hospital (KR)
Ministry of Health and Welfare, Ministry of Science and ICT, South Korea
Computational Drug Discovery Methods
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon