EconPapers    
Economics at your fingertips  
 

Knowledge-aware contrastive heterogeneous molecular graph learning

Mukun Chen, Jia Wu, Shirui Pan, Fu Lin, Bo Du, Xiuwen Gong and Wenbin Hu

PLOS Computational Biology, 2025, vol. 21, issue 5, 1-24

Abstract: Molecular representation learning is pivotal in predicting molecular properties and advancing drug design. Traditional methodologies, which predominantly rely on homogeneous graph encoding, are limited by their inability to integrate external knowledge and represent molecular structures across different levels of granularity. To address these limitations, we propose a paradigm shift by encoding molecular graphs into heterogeneous structures, introducing a novel framework: Knowledge-aware Contrastive Heterogeneous Molecular Graph Learning. This approach leverages contrastive learning to enrich molecular representations with embedded external knowledge. KCHML conceptualizes molecules through three distinct graph views—molecular, elemental, and pharmacological—enhanced by heterogeneous molecular graphs and a dual message-passing mechanism. This design offers a comprehensive representation for property prediction, as well as for downstream tasks such as drug-drug interaction prediction. Extensive benchmarking demonstrates KCHML’s superiority over state-of-the-art molecular property prediction models, underscoring its ability to capture intricate molecular features.Author summary: In the field of drug discovery, predicting molecular properties and drug interactions is crucial for developing new medications and ensuring patient safety. Traditional methods for representing molecular structures often fail to incorporate external knowledge and struggle to capture complex interactions at different levels of detail. To address these limitations, we developed a new framework called Knowledge-aware Contrastive Heterogeneous Molecular Graph Learning (KCHML). Our approach integrates information from three perspectives—molecular structure, elemental relationships, and pharmacological data—using advanced machine learning techniques. This combination allows for a more detailed and accurate representation of molecules, leading to better predictions of molecular properties and drug interactions. By improving how we model and understand molecules, our work has the potential to streamline drug development and reduce the risk of harmful drug interactions, contributing to safer and more effective treatments.

Date: 2025
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1013008 (text/html)
https://journals.plos.org/ploscompbiol/article/fil ... 13008&type=printable (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:plo:pcbi00:1013008

DOI: 10.1371/journal.pcbi.1013008

Access Statistics for this article

More articles in PLOS Computational Biology from Public Library of Science
Bibliographic data for series maintained by ploscompbiol ().

 
Page updated 2025-05-31
Handle: RePEc:plo:pcbi00:1013008