Characterize the difference between TMPRSS2-ERG and non-TMPRSS2-ERG fusion patients by clinical and biological characteristics in prostate cancer.

Abstract The TMPRSS2-ERG gene fusion were frequently found in prostate cancer, and thought to play some fundamental mechanisms for the development of prostate cancer. However, until now, the clinical and prognostic significance of TMPRSS2-ERG gene fusion was not fully understood. In this study, based on the 281 prostate cancers that constructed from a historical watchful waiting cohort, the statistically significant associations between TMPRSS2-ERG gene fusion and clinicopathologic characteristics were identified. In addition, the Elastic Net algorithm was used to predict the patients with TMPRSS2-ERG fusion status, and good predictive results were obtained, indicating that this algorithm was suitable to this prediction problem. The differential gene network was constructed from the network, and the KEGG enrichment analysis demonstrated that the module genes were significantly enriched in several important pathways. Highlights • The clinicopathologic differences between ERG positive and ERG negative patients were analyzed. • The Elastic Net was used to predict the TMPRSS2-ERG fusion status. • A network was constructed from the differential genes between two tumor groups. [ABSTRACT FROM AUTHOR]