سامانه همایش های دانشگاه آزاد اسلامی واحد نجف آباد

صفحه اصلی / سومین کنفرانس بین المللی مهندسی و علوم کامپیوتر و اولین کنفرانس ملی سامانه های هوشمند، کلان داده و رایانش پیشرفته

A Network-Based Computational Framework for Identifying Hub Genes in Polycystic Ovary Syndrome Using PPI Analysis

نویسندگان :

Reza Fani Kisomi¹ Farhad Mashayekhi²

1- دانشگاه گیلان 2- دانشگاه گیلان

کلمات کلیدی :

Polycystic Ovary Syndrome،Protein-Protein Interaction Network،Hub Gene Identification،Topological Clustering،Disease-Aware Scoring،MCODE

چکیده :

Polycystic Ovary Syndrome (PCOS) is a complex multifactorial disorder driven by interactions between genetic, molecular, and environmental factors. Traditional single-gene approaches fail to capture its underlying systems-level mechanisms. This study presents an integrated network-based computational framework for identifying and prioritizing hub genes in PCOS through PPI network analysis, topological clustering, and disease-aware composite scoring. Two hundred PCOS-associated proteins were retrieved from STRING database and constructed into a PPI network comprising 200 nodes and 4,979 edges. MCODE clustering identified eight dense modules, with Cluster 5 (17 proteins, 47 interactions) selected for detailed analysis. A composite scoring system integrating disease association scores (STRING), mean interaction confidence, and network topology (MCODE) ranked the top five hub genes as ERBB3 (3.338), INSR (3.298), NPY (3.234), PIK3CA (3.204), and AKT1 (3.060). These genes implicate insulin signaling, growth factor pathways, and neuroendocrine regulation in PCOS pathogenesis. The framework successfully distinguishes structurally central nodes from biologically relevant disease genes; notably, LPL—despite highest MCODE score (27.87)—ranked 8th due to low disease association (2.075). This balanced approach prevents over-identification of highly connected but disease-irrelevant hubs and under-weighting of biologically significant nodes, providing superior interpretability of PCOS molecular mechanisms. The methodology is generalizable to other complex polygenic diseases.