Identification of key genes in prostate cancer gene expression profile by bioinformatics

The aim of this study was to identify key candidate genes in prostate cancer. The gene expression profiles of GSE32448, GSE45016, GSE46602 and GSE104749 were downloaded from the Gene Expression Omnibus database. The differentially expressed genes (DEGs) between prostate cancer and normal samples were identified by R language. The gene ontology functional and pathway enrichment analyses of DEGs were performed by the Database for Annotation, Visualization and Integrated Discovery software followed by the construction of protein-protein interaction network. Hub gene identification was performed by the plug-in cytoHubba in Cytoscape software. The 217 DEGs were significantly enriched in biological processes including epithelial cell differentiation, response to estradiol and several pathways, mainly associated with protein digestion and absorption pathway in prostate cancer. Epithelial cell adhesion molecule, twist family basic helix-loop-helix transcription factor 1, CD38 molecule and vascular endothelial growth factor A were identified as hub genes. The expression levels of hub genes were consistent with data obtained in The Cancer Genome Atlas for prostate adenocarcinoma. These hub genes may be used as potential targets for prostate cancer diagnosis and treatment.