HetFCM: functional co-module discovery by heterogeneous network co-clustering
Nucleic Acids Research · 2024
DOI: 10.1093/nar/gkad1174 ↗Kurz gefasst
Fachartikel in der Zeitschrift Nucleic Acids Research, 2024. Verfasst von Haojiang Tan und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Software, Shandong University, Jinan 250101, Shandong, China. Hauptthemen laut MEDLINE: Breast Neoplasms, Cluster Analysis, Gene Regulatory Networks, Zea mays. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 100 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Major Research Plan, Fundamental Research Funds of Shandong University.
Abstract (Originalsprache)
Functional molecular module (i.e., gene–miRNA co-modules and gene–miRNA–lncRNA triple-layer modules) analysis can dissect complex regulations underlying etiology or phenotypes. However, current module detection methods lack an appropriate usage and effective model of multi-omics data and cross-layer regulations of heterogeneous molecules, causing the loss of critical genetic information and corrupting the detection performance. In this study, we propose a heterogeneous network co-clustering framework (HetFCM) to detect functional co-modules. HetFCM introduces an attributed heterogeneous network to jointly model interplays and multi-type attributes of different molecules, and applies multiple variational graph autoencoders on the network to generate cross-layer association matrices, then it performs adaptive weighted co-clustering on association matrices and attribute data to identify co-modules of heterogeneous molecules. Empirical study on Human and Maize datasets reveals that HetFCM can find out co-modules characterized with denser topology and more significant functions, which are associated with human breast cancer (subtypes) and maize phenotypes (i.e., lipid storage, drought tolerance and oil content). HetFCM is a useful tool to detect co-modules and can be applied to multi-layer functional modules, yielding novel insights for analyzing molecular mechanisms. We also developed a user-friendly module detection and analysis tool and shared it at http://www.sdu-idea.cn/FMDTool.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Substanzen
Autor:innen und Institutionen
- School of Software, Shandong University, Jinan 250101, Shandong, China.
- Joint SDU-NTU Centre for Artificial Intelligence Research, Shandong University, Jinan 250101, Shandong, China.
- College of Electrical and Information Engineering, Beijing Uni. of Civil Eng. and Arch., Beijing 100044, China.
- College of Agronomy & Biotechnolog, China Agricultural University, Beijing 100193, China.
ORCID-Profile: Jun Wang, Guoxian Yu
Förderung
- Major Research Plan
- Fundamental Research Funds of Shandong University
Zitieren
APA
Tan, H., Guo, M., Chen, J., Wang, J., & Yu, G. (2024). HetFCM: functional co-module discovery by heterogeneous network co-clustering. Nucleic Acids Research. https://doi.org/10.1093/nar/gkad1174
Vancouver
Tan H, Guo M, Chen J, Wang J, Yu G. HetFCM: functional co-module discovery by heterogeneous network co-clustering. Nucleic Acids Research. 2024. doi:10.1093/nar/gkad1174
BibTeX
@article{Tan2024,
author = {Tan, Haojiang and Guo, Maozu and Chen, Jian and Wang, Jun and Yu, Guoxian},
title = {HetFCM: functional co-module discovery by heterogeneous network co-clustering},
journal = {Nucleic Acids Research},
year = {2024},
doi = {10.1093/nar/gkad1174}
}
EUVIMED