Data Mining of Differentially Expressed Genes in Ovarian Epithelial Carcinoma: Implications in Precise Medicine

Abstract

Background: Epithelial Ovarian Carcinoma (EOC) is a ubiquitous gynecological malignancy with complicated pathogenesis. Genetic risk factors and pathways involved in the prognosis of this cancer are not yet understood completely. Determining genetic markers with diagnostic and prognostic values would pave the way for efficient management of cancer.

Objective: This study aimed to investigate the genes and the regulatory networks involved in the occurrence and prognosis of EOC through different bioinformatics analysis tools. In addition, recent advances in using bioinformatic analysis approaches based on the genes and regulatory networks, particularly Differentially Expressed Genes (DEGs) in improving the diagnosis and prognosis of EOC, are discussed.

Methods: The gene expression profiles of GSE18520, GSE54388, and GSE27651 were downloaded from the Gene Expression Omnibus (GEO) database and further analyzed with different analyses in R language. Current literature on using bioinformatics based on DEGs and associated regulatory networks to improve the diagnosis and prognosis of EOC was reviewed.

Results: Analyses of the gene expression levels between the malignant tissue against normal tissue unveiled 163 DEGs. Gene Ontology (GO) annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed on the target genes using cluster profile package and Cytoscape package was employed to assess the protein interaction network of these genes. The protein-protein interaction network was analyzed using the CytoHubba plug-in to identify 20 hub genes. In addition, we analyzed the prognosis of the hub genes using the Kaplan-Meier survival analysis that revealed evident differences in the prognosis of 13 genes. The malignant tissues exhibited a differential expression of 12 genes against healthy tissues as shown by Gene Expression Profiling Interactive Analysis (GEPIA) analysis.

Conclusion: Findings of this study revealed 12 genes to be significantly up-regulated and the prognosis was significantly different, which could be employed to potentially target EOC in clinical practice.

Keywords: Bioinformatics analysis; Differential expression; Epithelial ovarian cancer; GO.; Gene expression omnibus (GEO); Genes; Regulatory network.