Screening miRNA biomarkers for postmenopausal osteoporosis based on bioinformatics methods

Objective Using bioinformatics methods to screen potential miRNAs as biomarkers for postmenopausal osteoporosis (PMO). Methods Obtain the expression profile of PMO peripheral blood miRNA chip through the GEO public database. Firstly, the chip was re-annotated using R language, and then the clinical typing significance of the data was determined using similarity analysis (ANOSIM). Then, weighted gene co expression network analysis (WGCNA), multi-scale embedded gene co expression network analysis (MEGCNA), and nonnegative matrix factorization (NMF) were used to screen miRNAs related to PMO. Finally, the diagnostic efficacy of miRNA was evaluated using ROC curves, the target genes of miRNA were predicted using a database, and functional enrichment of the target genes was performed using Metascape.Results miR-223-3p has a high predictive diagnostic value for PMO. GO and KEGG enrichment analysis was conducted on 34 target genes potentially regulated by miR-223-3p, and the results showed that multiple pathways were associated with bone development.Conclusion miR-223-3p has significance in the diagnosis of PMO and may regulate bone development by regulating downstream target genes.