iProDNet: integrated probabilistic differential network inference under heterogeneous biological conditions
Briefings in Bioinformatics · 2026
DOI: 10.1093/bib/bbag565 ↗Kurz gefasst
Fachartikel in der Zeitschrift Briefings in Bioinformatics, 2026. Verfasst von Heewon Park und 1 weiteren Autor:innen. Beteiligte Einrichtung: School of Mathematics, Statistics and Data Science, Sungshin Women's University, 2, 34 dagil, Bomun-ro, Seongbuk-gu, Seoul, 02844, Republic of Korea. Hauptthemen laut MEDLINE: Gene Regulatory Networks, COVID-19, Computational Biology, SARS-CoV-2. Die Arbeit nennt 45 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: NRF.
Abstract (Originalsprache)
Understanding phenotype-specific network rewiring is essential for elucidating the molecular mechanisms underlying complex diseases. Although numerous differential network analysis methods have been developed, most focus on only a limited aspect of network information. Consequently, these approaches often fail to jointly characterize regulatory activity and network architecture, provide limited statistical evidence at the network level, and offer little capability for prioritizing key genes that drive phenotype-specific network rewiring. To address these limitations, we developed a novel computational framework, integrated probabilistic differential network analysis (iProDNet), for identifying phenotype-specific regulatory modules under heterogeneous biological conditions. A key feature of iProDNet is the integration of regulatory effects and network topological characteristics into a unified network-aware gene activity score. The framework further transforms gene-level differential activity into network-level statistical evidence using a nonparametric log-likelihood ratio approach and naturally prioritizes genes that contribute to phenotype-specific network rewiring. We evaluated iProDNet through Monte Carlo simulations and compared its performance with existing methods. Across diverse simulation settings, iProDNet consistently achieved superior discriminative performance while maintaining a balanced tradeoff between sensitivity and specificity. We further applied iProDNet to whole-blood RNA-sequencing data generated by the Japan Corona virus disease 2019 (COVID19) Task Force to identify molecular interactions associated with severe COVID19. The proposed framework identified distinct rewired subnetworks related to immune regulation, ribosome-mediated translation, and interferon-driven antiviral responses. Furthermore, iProDNet successfully prioritized biologically relevant markers, including B2M, ISG15, and multiple ribosomal protein genes, all of which have previously been implicated in COVID19 severity and antiviral immunity. These results demonstrate that iProDNet provides a statistically robust and biologically interpretable framework for identifying phenotype-specific regulatory modules and key regulatory drivers. As such, it offers new opportunities for investigating molecular rewiring mechanisms and discovering disease-associated therapeutic targets.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Mathematics, Statistics and Data Science, Sungshin Women's University, 2, 34 dagil, Bomun-ro, Seongbuk-gu, Seoul, 02844, Republic of Korea.
- Data Science Center, Sungshin Women's University, 2, 34 dagil, Bomun-ro, Seongbuk-gu, Seoul, 02844, Republic of Korea.
- M&D Data Science Center, Institute of Integrated Research, Institute of Science Tokyo, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8510, Japan.
- Human Genome Center, Institute of Medical Science, University of Tokyo, 4-6-1 Shirokanedai, Minato-ku, Tokyo, 108-8639, Japan.
ORCID-Profile: Heewon Park, Seiya Imoto
Förderung
- NRF
Zitieren
APA
Park, H., & Imoto, S. (2026). iProDNet: integrated probabilistic differential network inference under heterogeneous biological conditions. Briefings in Bioinformatics. https://doi.org/10.1093/bib/bbag565
Vancouver
Park H, Imoto S. iProDNet: integrated probabilistic differential network inference under heterogeneous biological conditions. Briefings in Bioinformatics. 2026. doi:10.1093/bib/bbag565
BibTeX
@article{Park2026,
author = {Park, Heewon and Imoto, Seiya},
title = {iProDNet: integrated probabilistic differential network inference under heterogeneous biological conditions},
journal = {Briefings in Bioinformatics},
year = {2026},
doi = {10.1093/bib/bbag565}
}
EUVIMED