Differential expression analysis in epithelial ovarian cancer using functional genomics and integrated bioinformatics approaches

Abstract

BackgroundEpithelial Ovarian Cancer (EOC) has remained the most frequent and leading cause of death among gynecologic malignancies, with a remarkably elevated and alarming global mortality rate and a poor prognosis. As a result of its asymptomatic characteristics in the early stages, the disease is typically diagnosed at advanced stages, with extensive dissemination. Therefore, it is of great importance to explore more reliable diagnostic and prognostic biomarkers. The present study intended to design an integrative bioinformatics approach to investigate robust differentially expressed genes (DEGs) associated with EOC progression as valuable diagnostic and prognostic biomarkers, providing inspiring insights into cancer mechanisms. Materials and methodsThree mRNA (GSE40595, GSE14407, and GSE18520) expression profiles related to EOC were retrieved from Gene Expression Omnibus (GEO) database. Significant DEGs were screened out following raw data quality assessment, preprocessing, and statistical computing. Gene ontology and pathway enrichment analyses were performed on candidate DEGs. Next, the interactions between genes were visualized by constructing their networks, and subsequently, a list of hub genes was extracted through the estimation of their connections, which were subjected to further evaluations for validation and survival analysis. ResultsWe identified 241 overlapping DEGs (20 up- and 221 down-regulated) between 3 mRNA datasets. After evaluating the protein-protein interactions, validation, and survival analysis, five hub genes including four up-regulated (AURKA, CD24, CDCA3, CENPF) and one down-regulated (PGR) were eventually selected as potential biomarkers of EOC. GO and Enriched pathways related to selected genes compromised cell growth, actin filament organization, enzyme inhibitor, peptidase regulator, and receptor protein kinase activities, glycosaminoglycan binding, glycine, serine, and threonine metabolism pathways, proteoglycans, and EGFR tyrosine kinase inhibitor resistance. ConclusionIn summary, these findings will shed further light on the molecular mechanisms underlying EOC. In addition, the candidate genes, related metabolites, and signaling pathways could serve as promising prognostic, diagnostic, or even potential drug targets.