Ovarian cancer (OC) is second most common form of gynaecological cancer world wide . In this study, we collected and analyzed three ovarian cancer microarray raw datasets from Gene Expression Omnibus, NCBI, and identified a total of 1806 significant DEGs (Differentially expressed genes). The functional analysis of the DEGs showed that the 885 upregulated DEGs were mostly enriched in protein-binding activity, while the downregulated 796 genes were mostly enriched in retinal dehydrogenase activity and GABA receptor binding. We then constructed a protein-protein interaction network of the DEGs DEGs in ovarian cancer datasetsand analyzed the network to find cluster subnets, using molecular complex detection (MCODE). Common genes among top hub gene list, bottleneck gene list and maximum clique centrality (MCC) gene lists were identified as key driver genes, After analyzing the network. The following genes, STK12 (Serine threonine protein kinase), UBE2C (Ubiquitin-conjugating enzyme E2 C), CENPA (Centromere protein A), CCNB1 (Cyclin B1), POLD1 (polymerase delta 1) and KIF11 (Kinesin Family Member 11) were finally identified as driver genes. Higher expression of the key driver genes, STK12, UBE2C, CENPA, CCNB1, POLD1 and KIF11, was associated with lower overall survival (OS) among ovarian cancer patients. Therefore, the identified driver genes could be important diagnostic and prognostic biomarkers for predicting ovarian cancer progression and understanding the mechanism of tumour formation and recurrence.

Keywords: Ovarian cancer; bottleneck genes; maximum clique centrality; network analysis.