Unified and explainable molecular representation learning for imperfectly annotated data from the hypergraph view

Bowen Wang, Junyou Li, Donghao Zhou, Lanqing Li, Jinpeng Li, Ercheng Wang, Jianye Hao (), Liang Shi, Chengqiang Lu, Jiezhong Qiu, Tingjun Hou (), Dongsheng Cao (), Guangyong Chen () and Pheng Ann Heng
Additional contact information
Bowen Wang: The Chinese University of Hong Kong
Junyou Li: Zhejiang Lab
Donghao Zhou: The Chinese University of Hong Kong
Lanqing Li: The Chinese University of Hong Kong
Jinpeng Li: The Chinese University of Hong Kong
Ercheng Wang: Zhejiang Lab
Jianye Hao: Tianjin University
Liang Shi: University of California
Chengqiang Lu: University of Science and Technology of China
Jiezhong Qiu: Zhejiang Lab
Tingjun Hou: Zhejiang University
Dongsheng Cao: Central South University
Guangyong Chen: Chinese Academy of Science
Pheng Ann Heng: The Chinese University of Hong Kong

Nature Communications, 2025, vol. 16, issue 1, 1-18

Abstract: Abstract Molecular representation learning (MRL) has shown promise in accelerating drug development by predicting chemical properties. However, imperfectly annotation among datasets pose challenges in model design and explainability. In this work, we formulate molecules and corresponding properties as a hypergraph, extracting three key relationships: among properties, molecule-to-property, and among molecules, and developed a unified and explainable multi-task MRL framework, OmniMol. It integrates a task-related meta-information encoder and a task-routed mixture of experts (t-MoE) backbone to capture correlations among properties and produce task-adaptive outputs. To capture underlying physical principles among molecules, we implement an innovative SE(3)-encoder for physical symmetry, applying equilibrium conformation supervision, recursive geometry updates, and scale-invariant message passing to facilitate learning-based conformational relaxation. OmniMol achieves state-of-the-art performance in properties prediction, reaches top performance in chirality-aware tasks, demonstrates explainability for all three relations, and shows effective performance in practical applications. Our code is available in our https://github.com/bowenwang77/OmniMol public repository.

Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.nature.com/articles/s41467-025-63730-6 Abstract (text/html)

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:nat:natcom:v:16:y:2025:i:1:d:10.1038_s41467-025-63730-6

Ordering information: This journal article can be ordered from
https://www.nature.com/ncomms/

DOI: 10.1038/s41467-025-63730-6

Access Statistics for this article

Nature Communications is currently edited by Nathalie Le Bot, Enda Bergin and Fiona Gillespie

More articles in Nature Communications from Nature
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().