【Objective】 The two databases，GEO（gene expression omnibus，GEO）and TCGA（the cancer genome alas ，TCGA），were analyzed using bioinformatics methods to screen differentially expressed genes associated and their related regulatory networks in prostate carcinoma. 【Methods】 The prostate carcinoma gene expression chip data （GSE46602 ，GSE55945） downloaded from the GEO database were integrated into the RNA- seq data of the TCGA database. And the differentially expressed genes analysis was performed using GEO2R and the edgeR package of R software to extract common significant differentially expressed genes. The clusterProfiler package of R software was used to enrich the GO（gene ontology ，GO）function enrichment analysis and KEGG（kyoto encyclopedia of genes and genomes， KEGG）pathway analysis. Differentially expressed genes were further constructed into a protein-protein interaction（PPI） network to screen out key genes for regulatory protein expression in prostate carcinoma. Gene analysis results were combined with TCGA clinical follow-up data to analyze the clinical prognostic value of key node genes. 【Results】A total of 278 significant differentially expressed genes were extracted，of which 178 genes were down- regulated and 100 genes were up-regulated. These genes were closely associated with the function and pathway enrichment such as the regulation of proliferation of epithelial cells，metabolism of benzene- containing compounds，the glutathione metabolism，and focal adhesion. The protein-protein interaction network analysis revealed three key protein expression modules and 12 key node genes. Among these key genes，EDN3（endothelin-3），EDNRB（endothelin receptor B）and AMACR（alpha-methylacyl- coa racemase）were closely related to the survival rate of prostate cancer patients. 【Conclusion】Through bioinformatics analysis of gene chip and RNA-seq data in prostate carcinoma，we found that EDN3，EDNRB and AMACR may play an important role in the occurrence and development of prostate carcinoma.