Abstract
Atherosclerosis (AS) is a growing vascular disease linked to plaque buildup, causing blood flow issues. Current diagnosis relies on symptoms and imaging, which are limited for early detection and plaque biology assessment. Treatments focus on symptoms but don't address root causes, leading to complications. This study aims to find new diagnostic markers and therapies using bioinformatics and machine learning. Data from gene expression omnibus datasets (GSE28829 for gene expression, GSE159677 for single-cell analysis) were analyzed via WGCNA to identify gene modules, Limma for differentially expressed genes, and gene ontology/KEGG for pathway enrichment. Protein-Protein Interaction networks, machine learning (least absolute shrinkage and selection operator, Random Forest, artificial neural network), immune infiltration (CIBERSORT), and single-cell RNA-seq were used. A nomogram model was built, and candidate drugs (e.g., simvastatin) were tested via molecular docking. Key modules (turquoise) and 238 differentially expressed genes linked to immune processes. Four biomarkers (toll like receptor 2, CCR1, interferon regulatory factor 8, CCL4) showed high diagnostic accuracy (AUC > 0.8). Immune analysis revealed altered macrophage/T cell profiles, with biomarkers correlating to monocyte/macrophage activity. The nomogram model was robust, and simvastatin docked strongly to target proteins. toll like receptor 2, CCR1, interferon regulatory factor 8, and CCL4 are novel AS biomarkers linked to immune pathways. The nomogram aids risk prediction, and simvastatin shows potential as a targeted therapy. Findings advance AS understanding and offer tools for early diagnosis and personalized treatment.