Abstract :Prostate Cancer (PCa) is one of the leading causes of cancer-related deaths among the men around
the world. In this study, we aim to identify candidate biomarkers in PCa using bioinformatics
analysis combined with the analysis of the common database of tumors and uncover possible
mechanisms. The gene expression profiles of GSE55945 including 13 PCa samples (with Gleason
score of 6 or 7) and 8 normal prostate samples were downloaded from GEO database. Firstly,
Differentially Expressed Genes (DEGs) were obtained using “limma” R package followed by preprocession
of raw expression data. A total of 581 genes, including 204 up-regulated genes and 377
down-regulated genes, were screened out in PCa tissues compared with normal prostate tissues
with the cut-off criteria p<0.05 and |log2FC|>1. Secondly, the Gene Ontology (GO) and Kyoto
Encyclopedia of Genes and Genomes pathway (KEGG) enrichment analyses were performed using
DAVID database. Thirdly, protein-protein interaction (PPI) network of the DEGs was constructed
by Cytoscape software. Modules in PPI network were screened using Molecular Complex Detection
(MCODE). At last, 7 hub genes, ANXA1, CHRM3, UTS2, PROK1, AGT, CCK and EDN3 were
identified from the modules of PPI network, and then validated by Oncomine database and Protein
atlas database. In conclusion, our study suggested that the identified DEGs and hub genes promote
our understanding of the molecular mechanisms underlying the development of PCa, and might
reveal preliminary information with regard to carcinogenesis of prostate cancer.