Combining multi-dimensional data to identify a key signature (gene and miRNA) of cisplatin-resistant gastric cancer

Gastric cancer (GC) is one of the most lethal malignant tumors; the resistance of this type of tumor is the main source of GC treatment failure. In this study, we used bioinformatics analysis to verify differences in resistant GCs and identify an effective method for reversing drug resistance in GC. Microarray data [gene and microRNA (miRNA)] were analyzed using GEO2R software, and Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were applied to further enrich the genetic data. miRNA-gene interactions were determined using Cytoscape (v.3.5.1). Online software was used to analyze protein interactions and predict network structure. The Cancer Genome Atlas (TCGA) database was used to verify the expression levels of genes in GC resistance. miR-604 expression levels were verified by real-time PCR in GC cell lines. We screened 3981 GC resistance-associated genes and 244 miRNAs using bioinformatics methods. Six hub genes were identified and verified in the TCGA database, including five up-regulated genes, POLR2L, POLR2C, POLR2F, APRT, and LMAN2, and a down-regulated gene, NFKB2. The up-regulated genes POLR2L, POLR2C, APRT, and LMAN2 interact with miR-604; therefore, we focused on miR-604, which has low expression in drug-resistant GC. The results of this study indicate that through bioinformatics technologies, we have determined the hub genes and hub miRNAs related to drug resistance in GC. Among them, miR-604 could become a new indicator in the diagnosis of drug-resistant GC and may be used to investigate the pathogenesis of resistance in GC.