Employing an integrated bioinformatics and systems biology approach to uncover key genes and drug targets for ovarian cancer

Abstract

Ovarian cancer (OC) is a common disease in females lacking precise diagnostic and therapeutic biomarkers. The access to high-yield genomic data and availability of advanced Bioinformatics tools permit to envisage disease sensitive biomarkers and their drug targets that could be re-modeled for better patient outcomes. In this study, we analyzed three GEO datasets: GSE54388, GSE36668, and GSE18520 including 20 normal and 77 OC samples using GEO2R to identify differentially expressed genes (DEGs) in OC. Functional annotation and KEGG pathway analysis of DEGs were performed using DAVID and FunRich tools. Protein-protein interactions were extracted via GeneMANIA and analyzed using Cytoscape. We identified 115 DEGs, including 84 upregulated and 31 downregulated genes. Key biological processes of OC DEGs were signal transduction, cell growth and maintenance, cell communication, cell-cell adhesion, and apoptosis. Enriched KEGG pathways (P-value<0.05) included focal adhesion, P13K-Akt, JAK-STAT, Wnt-signaling, and pathways in cancer. Module analysis identified six seed/hub genes, with five genes viz. PDGFRA, DIRAS3, HHIP, PROK1, and MTUS1 showed significant impact on the overall survival of OC patients (P-value <0.05). Additionally, DIRAS3, and MTUS1 were found to be hypermethylated in ovarian cancer (OC). Treatment with Valproic acid (2-propylpentanoic acid) reactivates these genes by reversing their epigenetic silencing. These results suggest that DIRAS3, and MTUS1 may serve as potential OC biomarkers, while Valproic acid could be a promising therapeutic agent for targeting their epigenetic modifications in OC.