Recent studies have shown that dysfunctional microRNAs (miRNAs) are involved in the progression of various cancers. Dysfunctional miRNAs may jointly regulate their target genes and further alter the activities of canonical biological pathways. Identification of the pathways regulated by a group of dysfunctional miRNAs could help uncover the pathogenic mechanisms of cancer and facilitate development of new drug targets. Current miRNA-pathway analyses mainly use differentially-expressed miRNAs to predict the shared pathways on which they act. However, these methods fail to consider the level of differential expression level, which could improve our understanding of miRNA function. We propose a novel computational method, MicroRNA Set Enrichment Analysis (MiRSEA), to identify the pathways regulated by dysfunctional miRNAs. MiRSEA integrates the differential expression levels of miRNAs with the strength of miRNA pathway associations to perform direct enrichment analysis using miRNA expression data. We describe the MiRSEA methodology and illustrate its effectiveness through analysis of data from hepatocellular cancer, gastric cancer and lung cancer. With these analyses, we show that MiRSEA can successfully detect latent biological pathways regulated by dysfunctional miRNAs. We have implemented MiRSEA as a freely available R-based package on CRAN (https://cran.r-project.org/web/packages/MiRSEA/).
Keywords: cancer; enrichment analysis; mRNA; microRNA; pathway.