Identification of drug targets towards prostate canser

Abstract

Prostate cancer (Pca) is one of the leading cause of cancer deaths and second common cancer type in men. Identifying novel diagnostics and drug targets specific for Pca is critical since early detection and treatment are associated with high sur vival rates. As intracellular processes are interconnected, a systematic point of view is essential instead of analysis of individual components. Biological networks such as protein-protein interaction (PPI) networks have been proven effective representations of biological data. In this study, two types of networks were constructed and ana lyzed: Protein-protein interaction network (PPIN) was constructed by a selecting core set of Pca proteins and combining experimental interactions obtained from 4 differ ent databases by core set proteins’ mutual Gene Ontology (GO) terms. Co-Expression (Co-Exp) network was obtained by Weighted Correlation Network Analysis (WGCNA) using transcriptomics data from The Cancer Genome Atlas (TCGA). Both networks were divided into modules by Markov Clustering (MCL) and Topological Overlap Ma trix (TOM), respectively. Gene Set Enrichment Analysis (GSEA) demonstrated that prostate tumors cover the eight hallmarks of cancer. Disease Ontology (DO) enrich ments and disease identification of genes detected by the DIAMOnD algorithm revealed the link between Pca and other cancer types such as sarcoma, liver and bone cancer. Integration of topological features and differential expression of genes led to the iden tification of 272 potential drug targets. Further elimination of these targets was done by screening the Connectivity Map (C-Map) and 84 potential targets were identified, which are proteins taking part in hallmarks of cancer processes. 55 of these promising spots without any known drugs need further computational and experimental exami nation to be considered as new putative targets.