Abstract

Osteoporosis (OP) is a prevalent chronic systemic metabolic bone disease, with its pathogenesis remaining largely unclear. This study employs machine learning and bioinformatics analyses to identify the hub genes associated with OP and to investigate the role of immune phenotype shifts in this condition. Microarray technology was utilized to analyze gene expression profiles in both OP and control group, with data integration sourced from the GEO database. Differentially expressed genes were identified using the Limma method. The biological functions of these genes were examined through Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, Gene Ontology (GO) functional analysis, and Gene Set Enrichment Analysis (GSEA). Additionally, Machine learning was employed to filter out the hub genes of OP, and Receiver Operating Characteristic (ROC) analysis was utilized to assess the diagnostic performance of these hub genes. Validation of hub genes was performed using quantitative Reverse Transcription-Polymerase Chain Reaction (qRT-PCR) on peripheral blood sample obtained from OP and control group. Finally, immune cell composition analysis was conducted to evaluate immune cell dysregulation in OP, concurrently exploring the correlation between immune cell and hub genes. We identified 12 differentially expressed genes in OP. Machine learning algorithms further identified CCR5 and KIF13B as hub genes. ROC curves demonstrated the high diagnostic efficacy of these genes. The results of qRT-PCR showed that the expression levels of hub genes were consistent with those observed in the microarray analysis. Additionally, immune cell composition analysis suggested a close correlation between OP and imbalances in NK cell levels, with CCR5 showing a negative correlation with mast cells. Two hub genes associated with OP, namely CCR5 and KIF13B, have been identified a