Abstract
PURPOSE: Postmenopausal osteoporosis (PMOP) is a common and debilitating chronic disease, but it has just no cure options. The objective of this study was to identify genes associated with osteoporosis and reveal potential therapeutic targets. METHODS: Expression profiles from GSE13850 and GSE56815 datasets were combined for differential expression analysis. Extraction of intersecting genes from the combined datasets and the differentially expressed genes in GSE56814 were performed to construct a multi-scale embedded gene co-expression network analysis (MEGENA) to obtain module genes. Module genes with an area under the receiver operating characteristic curve (AUC) >0.60 were chosen to construct the least absolute shrinkage and selection operator (LASSO) model to obtain feature genes. A regulated network was constructed using differentially expressed micro-RNAs (miRNAs) in GSE74209 and feature genes. Finally, key genetic pathways and pathways of the Kyoto Encyclopedia of Genes and Genomes were identified and explored. RESULTS: The commonly identified differentially expressed genes involve oxidative phosphorylation and caffeine metabolism. We identified 66 modules with 2354 module genes based on MEGENA. CARD8, FOXO4, IL1R2, MPHOSPH6, MPRIP, MYOM1, PRR5L and YIPF4 were identified as feature genes by the LASSO model. Furthermore, predicted miRNA target genes included 8 genes associated with PMOP. The largest AUC was observed for FOXO4, which was found at the nexus of feature genes and miRNA-regulated genes and which correlated with the upregulation of dendritic cells. Moreover, FOXO4 was found to be involved in ABC transporters, as well as cocaine and nicotine addiction. CONCLUSION: FOXO4 may serve as potential biomarker and therapeutic target for PMOP.