Identifying Network Biomarkers in Early Diagnosis of Hepatocellular Carcinoma via miRNA–Gene Interaction Network Analysis

Background: Hepatocellular carcinoma (HCC) is a highly heterogeneous cancer at the histological level. Despite the emergence of new biological technology, advanced-stage HCC remains largely incurable. The prediction of a cancer biomarker is a key problem for targeted therapy in the disease. Methods: We performed a miRNA–gene integrated analysis to identify differentially expressed miRNAs (DEMs) and genes (DEGs) of HCC. The DEM–DEG interaction network was constructed and analyzed. Gene ontology enrichment and survival analyses were also performed in this study. Results: By the analysis of healthy and tumor samples, we found that 94 DEGs and 25 DEMs were significantly differentially expressed in different datasets. Gene ontology enrichment analysis showed that these 94 DEGs were significantly enriched in the term “Liver” with a statistical p-value of 1.71 × 10(−26). Function enrichment analysis indicated that these genes were significantly overrepresented in the term “monocarboxylic acid metabolic process” with a p-value = 2.94 × 10(−18). Two sets (fourteen genes and five miRNAs) were screened by a miRNA–gene integrated analysis of their interaction network. The statistical analysis of these molecules showed that five genes (CLEC4G, GLS2, H2AFZ, STMN1, TUBA1B) and two miRNAs (hsa-miR-326 and has-miR-331-5p) have significant effects on the survival prognosis of patients. Conclusion: We believe that our study could provide critical clinical biomarkers for the targeted therapy of HCC.