Potential diagnostic markers shared between non-alcoholic fatty liver disease and atherosclerosis determined by machine learning and bioinformatic analysis.

Background: Evidence indicates that chronic non-alcoholic fatty liver disease (NAFLD) can increase the risk of atherosclerosis (AS), but the underlying mechanism remains unclear. Objective: This study is intended for confirming key genes shared between NAFLD and AS, and their clinical diagnostic value to establish a foundation for searching novel therapeutic targets. Methods: We downloaded the Gene Expression Omnibus (GEO) datasets, GSE48452 and GSE89632 for NAFLD and GSE100927, GSE40231 and GSE28829 for AS. The progression of NAFLD co-expression gene modules were recognized via weighted gene co-expression network analysis (WGCNA). We screened for differentially expressed genes (DEGs) associated with AS and identified common genes associated with NAFLD and AS using Venn diagrams. We investigated the most significant core genes between NAFLD and AS using machine learning algorithms. We then constructed a diagnostic model by creating a nomogram and evaluating its performance using ROC curves. Furthermore, the CIBERSORT algorithm was utilized to explore the immune cell infiltration between the two diseases, and evaluate the relationship between diagnostic genes and immune cells. Results: The WGCNA findings associated 1,129 key genes with NAFLD, and the difference analysis results identified 625 DEGs in AS, and 47 genes that were common to both diseases. We screened the core Conclusion: We identified