Multi-Omics and Machine Learning Integration Identifies Key Endothelial Modules in Carotid Artery Stenosis

<h4>Background</h4>Carotid artery stenosis (CAS) is a major cause of ischemic stroke, yet reliable molecular biomarkers for early identification remain limited.<h4>Methods</h4>We integrated single-cell RNA sequencing (scRNA-seq), bulk transcriptomics, and in-house multi-omics data, applying WGCNA and machine learning to identify endothelial cell-derived diagnostic biomarkers, validated across independent GEO and ZZ cohorts at single-cell, transcriptomic, and proteomic levels.<h4>Results</h4>scRNA-seq identified endothelial enrichment in CAS and yielded 836 markers. Integrative analysis (DEGs + WGCNA) defined 80 candidates, with NRP1 and XAF1 selected by machine learning. Both were consistently upregulated and validated across multi-omics datasets, showing strong diagnostic performance.<h4>Conclusion</h4>NRP1 and XAF1 represent novel endothelial cell-derived biomarkers with potential utility for early CAS screening and clinical diagnosis.