Ke Cai, Dongjie Chen, Jingyu Wang, Jing Bai, Mengting Wang, Jianming Guo, Feng Liu, Jin-Ao Duan, Shulan Su
Many patients with atherosclerotic cardiovascular disease (ASCVD) do not respond well to aspirin therapy. This study employed a multi-omics approach to identify key genes associated with aspirin non-responsiveness, thereby providing potential molecular targets and theoretical basis for improving the precision treatment strategies for ASCVD patients. One aspirin non-responsiveness dataset and two atherosclerosis-related datasets were obtained from the Gene Expression Omnibus database. After identifying differentially expressed genes through weighted gene co-expression network analysis and differential expression analysis, we used machine learning techniques to screen for key genes and establish predictive models. The results showed that two key genes (CLEC7A and RGS1) performed well in diagnosing aspirin non-responsiveness and ASCVD (area under the curve = 0.701-0.986). Among them, CLEC7A was significantly overexpressed in the aspirin non-responsiveness and ASCVD datasets (P < 0.0001) and was identified as a potential risk factor. Furthermore, ELISA results showed overexpression of CLEC7A in the aortic tissue and serum of ASCVD mice (P < 0.0001). In addition, immune infiltration analysis suggested that aspirin non-responsiveness mediated by CLEC7A may involve natural killer cells (P = 0.04). However, CLEC7A-mediated ASCVD may involve regulatory T cells and neutrophils (P < 0.0001), a finding further supported by immunofluorescence co-localization analyses. We constructed a regulatory network for CLEC7A and conducted small-molecule virtual screening for CLEC7A. In summary, CLEC7A can serve as a potential biomarker to identify ASCVD patients with a low aspirin response.