Abstract
Cuproptosis, a copper-dependent form of cell death, has been implicated in immune function and osteoporosis. However, the specific roles of cuproptosis-related genes (CRGs) in osteoporosis remain unclear. The differentially expressed CRGs from the Gene Expression Omnibus datasets of persons with osteoporosis and healthy individuals were categorized using R software tools in this study. Following that, the CIBERSORT algorithm and the GSVA technique were used to investigate the relationships between the different clusters and immune infiltration characteristics. Based on four machine learning techniques (Random Forest, Support Vector Machine, XGBoost, and Generalized Linear Model), Support Vector Machine and WGCNA analysis was carried out to identify the main genes linked to cuproptosis in the pathological course of osteoporosis. Subsequently, a model was built using the core genes related to cuproptosis to forecast the disease and identify potential treatment targets. The model was validated using an external dataset. In the end, a nomogram and calibration curve were created to improve this model's clinical applicability. Additionally, to investigate the possible biological roles of the core genes related to cuproptosis, we enriched them along several pathways. This study represents the first identification of key CRGs and core genes associated with cuproptosis in osteoporosis patients, findings that will facilitate the development of novel therapeutic strategies.