Alzheimer’s disease (AD), a common neurological disorder, has no effective treatment due to its complex pathogenesis. Disulfidptosis, a newly discovered type of cell death, seems to be closely related to the occurrence of various diseases. In this study, through bioinformatics analysis, the expression and function of disulfidptosis-related genes (DRGs) in Alzheimer’s disease were explored.
Differential analysis was performed on the gene expression matrix of AD, and the intersection of differentially expressed genes and disulfidptosis-related genes in AD was obtained. Hub genes were further screened using multiple machine learning methods, and a predictive model was constructed. Finally, 97 AD samples were divided into two subgroups based on hub genes.
In this study, a total of 22 overlapping genes were identified, and 7 hub genes were further obtained through machine learning, including MYH9, IQGAP1, ACTN4, DSTN, ACTB, MYL6, and GYS1. Furthermore, the diagnostic capability was validated using external datasets and clinical samples. Based on these genes, a predictive model was constructed, with a large area under the curve (AUC = 0.8847), and the AUCs of the two external validation datasets were also higher than 0.7, indicating the high accuracy of the predictive model. Using unsupervised clustering based on hub genes, 97 AD samples were divided into Cluster1 (
A close association between disulfidptosis and Alzheimer’s disease was discovered in this study, and a predictive model was established to assess the risk of disulfidptosis subtype in AD patients. This study provides new perspectives for exploring biomarkers and potential therapeutic targets for Alzheimer’s disease.