Interpretable high-order knowledge graph neural network for predicting synthetic lethality in human cancers
Briefings in Bioinformatics · 2025
DOI: 10.1093/bib/bbaf142 ↗Kurz gefasst
Fachartikel in der Zeitschrift Briefings in Bioinformatics, 2025. Verfasst von Xuexin Chen und 5 weiteren Autor:innen. Beteiligte Einrichtung: School of Computer Science, Guangdong University of Technology, No. 100 Waihuan Xi Road, Panyu, Guangdong, Guangzhou, 510006, China. Hauptthemen laut MEDLINE: Neoplasms, Neural Networks, Computer, Synthetic Lethal Mutations, Computational Biology. Die Arbeit wurde bisher 10-mal zitiert (Crossref) und nennt 54 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Science and Technology Major Project, National Science Fund for Excellent Young Scholars, Natural Science Foundation of China.
Abstract (Originalsprache)
Synthetic lethality (SL) is a promising gene interaction for cancer therapy. Recent SL prediction methods integrate knowledge graphs (KGs) into graph neural networks (GNNs) and employ attention mechanisms to extract local subgraphs as explanations for target gene pairs. However, attention mechanisms often lack fidelity, typically generate a single explanation per gene pair, and fail to ensure trustworthy high-order structures in their explanations. To overcome these limitations, we propose Diverse Graph Information Bottleneck for Synthetic Lethality (DGIB4SL), a KG-based GNN that generates multiple faithful explanations for the same gene pair and effectively encodes high-order structures. Specifically, we introduce a novel DGIB objective, integrating a determinant point process constraint into the standard information bottleneck objective, and employ 13 motif-based adjacency matrices to capture high-order structures in gene representations. Experimental results show that DGIB4SL outperforms state-of-the-art baselines and provides multiple explanations for SL prediction, revealing diverse biological mechanisms underlying SL inference.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Computer Science, Guangdong University of Technology, No. 100 Waihuan Xi Road, Panyu, Guangdong, Guangzhou, 510006, China.
- Pazhou Laboratory (Huangpu), No. 248 Pazhou Qiaotou Street, Haizhu, Guangdong Province, Guangzhou, 510335, China.
- Machine Learning Department, Mohamed bin Zayed University of Artificial Intelligence, Masdar, Abu Dhabi, United Arab Emirates.
- School of Information Science and Technology, ShanghaiTech University, No. 393 Huaxia Middle Road, Pudong, Shanghai, 201210, China.
- School of Information Science and Technology, Shanghai Engineering Research Center of Intelligent Vision and Imaging, ShanghaiTech University, No. 393 Huaxia Middle Road, Pudong, Shanghai, 201210, China.
- Institute for Infocomm Research (I2R), A*STAR, No. 2 Fusionopolis Way, Queenstown Planning, Singapore 138632, Singapore.
ORCID-Profile: Xuexin Chen, Jie Zheng, Min Wu
Förderung
- National Science and Technology Major Project
- National Science Fund for Excellent Young Scholars
- Natural Science Foundation of China
- A*STAR's Decentralised Gap Funding
- A*STAR’s Decentralised Gap Funding
Zitieren
APA
Chen, X., Cai, R., Huang, Z., Li, Z., Zheng, J., & Wu, M. (2025). Interpretable high-order knowledge graph neural network for predicting synthetic lethality in human cancers. Briefings in Bioinformatics. https://doi.org/10.1093/bib/bbaf142
Vancouver
Chen X, Cai R, Huang Z, Li Z, Zheng J, Wu M. Interpretable high-order knowledge graph neural network for predicting synthetic lethality in human cancers. Briefings in Bioinformatics. 2025. doi:10.1093/bib/bbaf142
BibTeX
@article{Chen2025,
author = {Chen, Xuexin and Cai, Ruichu and Huang, Zhengting and Li, Zijian and Zheng, Jie and Wu, Min},
title = {Interpretable high-order knowledge graph neural network for predicting synthetic lethality in human cancers},
journal = {Briefings in Bioinformatics},
year = {2025},
doi = {10.1093/bib/bbaf142}
}
EUVIMED