The increasing advances in genomic technologies in the last decade have allowed us to understand the molecular mutational landscape of prostate cancer. However, validation of genomic profiles generated by high-throughput efforts is laborious and expensive. Therefore, there is a need for a systematic and streamlined assessment of high-throughput genomic data to prioritise genes for further detailed biological validation studies for which this thesis entailed. Through cluster analysis of a panel of carefully selected markers, such as AR, ERG, MYC, RB1, PTEN and TP53, we were able to align patients into individual subgroups based on their PTEN status. Subsequently, through an objective computer learning elastic net modelling, we identified a cluster of 35 genes that was defining the clusters in our cohort. The prognostic effect of this signature was conserved in three independent datasets, with prominent statistical power, in Gleason 7 prostate cancer. Notably, our study is the first to report a signature with prognostic value in Gleason 7 cases. Additionally, we were able to identify a putative actionable target, S1PR2, by overlapping PTEN-low expressing clinical cases with gene expression data available from a Pten-knockout mouse model. From our analysis, we have observed that the combined inhibition of S1PR2 and the S1P kinases, SPHK1/2 were able to reduce cell migration and viability. Through a series of molecular validation, we postulated that the inhibition of both S1PR2 and SPHK1/2 could decrease cell viability and migration essential for cancer cell survival. Collectively, our findings contributed to the better understanding of the genomic changes associated with the heterogeneity of prostate cancer. Furthermore, we have uncovered several possible avenues for novel therapeutic interventions for untreated Gleason 7 prostate cancer.