Identification of Metabolism-Associated Prostate Cancer Subtypes and Construction of a Prognostic Risk Model.

摘要: Background: Despite being the second most common tumor in men worldwide, metabolism-associated mechanisms of prostate cancer (PCa) remain unclear. Herein, this study aimed to investigate characteristics PCa and develop a metabolism- associated prognostic risk model for patients with PCa. Methods: The activity levels metabolic pathways were determined using mRNA expression profiling Cancer Genome Atlas Prostate Adenocarcinoma cohort via single-sample gene set enrichment analysis (ssGSEA). analyzed samples divided into three subtypes based on partitioning around medication algorithm. Tumor subsets then investigated t-distributed stochastic neighbor embedding (t-SNE) analysis, differential Kaplan-Meier survival GSEA. Finally, we developed validated weighted co-expression network univariate Cox least absolute shrinkage selection operator, multivariate analysis. Other cohorts (GSE54460, GSE70768, genotype-tissue expression, International Consortium) utilized external validation. Drug sensibility was performed Genomics Sensitivity GSE78220 datasets. In total, 1039 six cell lines concluded our work. Results: Three clusters significantly different disease-free (DFS), clinical stage, stemness index, microenvironment including stromal immune cells, DNA mutation (TP53 SPOP), copy number variation, microsatellite instability identified Eighty-four module genes narrowed six-gene signature DFS, CACNG4, SLC2A4, EPHX2, CA14, NUDT7, ADH5 (p < 0.05). A developed, validation revealed strong robustness possessed diagnosis prognosis as well association feature drug sensitivity. Conclusions: reflected pathogenesis, essential features, heterogeneity tumors. Our may provide clinicians predictive values diagnosis, prognosis, treatment guidance