Bioinformatics analyses of the pathogenesis and new biomarkers of chronic obstructive pulmonary disease.

Background: Chronic obstructive pulmonary disease (COPD) is one of the major cause of global death. The purpose of our analysis was to detect a more reliable biomarker and small-molecule drug candidates and to identify the precise mechanisms involved in COPD.

Methods: Three data sets were downloaded from the Gene Expression Omnibus database and analysed by Gene Expression Omnibus 2R. Functional enrichment analyses were performed by Metascape. We use the STRING data to build a protein-protein interaction network. The targets of differentially expressed microRNA (DE miRNA) were predicted by the miRWalk database. Small-molecule drugs were predicted on connectivity map.

Results: A total of 181 differentially expressed genes and 35 DE miRNAs were confirmed. The protein-protein interaction network including all integrated differentially expressed genes was constructed, and 4 modules were filtrated. The module genes were relative to immune, inflammatory and oxidative stress functions according to a pathway analysis. The top 20 key genes were screened. Among the DE miRNAs found to be regulating key genes, miR-194-3p, MiR-502-5p, MiR-5088-5p, MiR-3127-5p, and miR-23a-5p might be the most significant due to their high number of connecting nodes in COPD. In addition, cephaeline, emetine, gabapentin, and amrinone were found to be potential drugs to treat COPD patients.

Conclusion: Our study suggests that miR-194-3p, miR-502-5p, and miR-23a-5p might participate in the nosogenesis of COPD. In addition, 4 potential small-molecule drugs were considered potentially useful for treating COPD patients.