Systematic prediction of target genes and pathways in cervical cancer from microRNA expression data

Cervical cancer (CC) is a leading cause of cancer-associated mortality in women; thus, the present study aimed to investigated potential target genes and pathways in patients with CC by utilizing an ensemble method and pathway enrichment analysis. The ensemble method integrated a correlation method [Pearson's correlation coefficient (PCC)], a causal inference method (IDA) and a regression method [least absolute shrinkage and selection operator (Lasso)] using the Borda count election algorithm, forming the PCC, IDA and Lasso (PIL) method. Subsequently, the PIL method was validated to be a feasible approach to predict microRNA (miRNA) targets by comparing predicted miRNA targets against those from a confirmed database. Finally, Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis was conducted for target genes in the 1,000 most frequently predicted miRNA-mRNA interactions to determine target pathways. A total of 10 target genes were obtained that were predicted >5 times, including secreted frizzled-related protein 4, maternally expressed 3 and NIPA like domain containing 4. Additionally, a total of 17 target pathways were identified, of which cytokine-cytokine receptor interaction (P=8.91×10^-7) was the most significantly associated with CC of all pathways. In conclusion, the present study predicted target genes and pathways for patients with CC based on miRNA expression data, the PIL method and pathway analysis. The results of the present study may provide an insight into the pathological mechanisms underlying CC, and provide potential biomarkers for the diagnosis and treatment of this tumor type. However, these biomarkers have yet to be validated; these validations will be performed in future studies.