Multi-omics data integration approach identifies potential biomarkers for Prostate cancer

Abstract Prostate cancer (PCa) is one of the most common malignancies, and many studies have shown that PCa has a poor prognosis, which varies across different ethnicities. This variability is caused by genetic diversity. High-throughput omics technologies have identified and shed some light on the mechanisms of its progression and finding new biomarkers. Still, a systems biology approach is needed for a holistic molecular perspective. In this study, we applied a multi-omics approach to data analysis using different publicly available omics data sets from diverse populations to better understand the PCa disease etiology. Our study used multiple omic datasets, which included genomic, transcriptomic and metabolomic datasets, to identify drivers for PCa better. Individual omics datasets were analysed separately based on the standard pipeline for each dataset. Furthermore, we applied a novel multi-omics pathways algorithm to integrate all the individual omics datasets. This algorithm applies the p-values of enriched pathways from unique omics data types, which are then combined using the MiniMax statistic of the PathwayMultiomics tool to prioritise pathways dysregulated in the omics datasets. The single omics result indicated an association between up-regulated genes in RNA-Seq data and the metabolomics data. Glucose and pyruvate are the primary metabolites, and the associated pathways are glycolysis, gluconeogenesis, pyruvate kinase deficiency, and the Warburg effect pathway. From the interim result, the identified genes in RNA-Seq single omics analysis are linked with the significant pathways from the metabolomics analysis. The multi-omics pathway analysis will eventually enable the identification of biomarkers shared amongst these different omics datasets to ease prostate cancer prognosis.