Comprehensive multi-omics analysis reveals the molecular mechanism of prostate cancer recurrence

Abstract

Background Accurately detecting prostate cancer recurrence currently poses a challenge for clinicians. In addition, biochemical recurrence (BCR) is a crucial risk factor for clinical recurrence and metastasis. The understanding of genes involved in BCR and their mechanisms is limited. Therefore, this study aims to comprehensively explore the genes associated with BCR and their biological mechanisms in prostate cancer using bioinformatics techniques. Methods Data from 473 non-recurrence (n = 412) and recurrence (n = 61) samples, were obtained from the TCGA public database. The key genes between groups were identified using the Limma package. Mendelian Randomization (MR) was employed to screen for key genes, describing their eQTL-positive outcomes in causality. Relationships between key genes and immune infiltration, immune cells, drug sensitivity, and signaling pathways were analyzed. Further, the enrichment of transcriptome gene sets, prediction of transcription factors, and specific situations in single cells were evaluated. Results In all, 486 DEGs were found, comprising 380 upregulated and 106 downregulated genes. MR identified DENND4B, KCNK6, MPHOSPH6, SPNS1, SYTL3, and ZNF419 as pivotal genes. Multi-omics analysis suggested these genes as predictive and diagnostic markers for BCR. Conclusion This study identified prostate cancer recurrence-related DEGs and their functions using bioinformatics and MR analysis, offering significant clinical implications for accurate prediction and assessment of prostate cancer recurrence. It also provided effective targets for managing recurrent prostate cancer.