Integrative analysis of cancer driver genes in prostate adenocarcinoma

Abstract

Large‑scale genomics studies have identified recurrently mutated genes in the ETS gene family, including fusions and copy number variations (CNVs), which are involved in the development of prostate adenocarcinoma (PRAD). However, the aetiology of PRAD remains to be fully elucidated. In the present study, 333 driver genes were identified using four computational tools: OncodriveFM, OncodriveCLUST, iCAGES and DrGaP. In addition, 32 driver pathways were identified using DrGaP. SPOP, TP53, SPTA1, AHNAK, HMCN1, ATM, FOXA1, CSMD3, LRP1B and FREM2 were the 10 most recurrently mutated genes in PRAD. ITGAL, TAGAP, SIGLEC10, RAC2 and ITGA4 were the five hub genes in the yellow module that were associated with the number of positive lymph nodes. Hierarchical clustering analysis of the 20 driver genes with the most frequent CNVs revealed three clusters of patients with PRAD. Cluster 3 tumours exhibited significantly higher numbers of positive lymph nodes, higher Gleason scores, more advanced cancer stages and poorer prognosis than cluster 1 and 2 tumours. A total of 48 genes were significantly associated with the number of positive lymph nodes, Gleason scores and pathologic stage in patients with PRAD. The identified set of cancer genes and pathways sheds light on the tumorigenesis of PRAD and creates avenues for the development of prognostic biomarkers and driver gene-targeted therapies in PRAD.