Development of immune-derived molecular markers for coronary heart disease via multimachine learning.

Coronary artery disease (CAD) is a major cardiovascular disorder primarily caused by coronary atherosclerosis, and early, sensitive diagnosis remains a clinical challenge. In this study, we developed a novel diagnostic model for CAD and explored potential therapeutic agents. CAD-associated datasets were obtained from the Gene Expression Omnibus. Using multimachine learning algorithms, 32 CAD- and immune-related characteristic genes were identified, and the resulting diagnostic model demonstrated excellent diagnostic performance. Immune cell infiltration analysis suggested that CD8+ T cells and naive B cells were the principal immune cell populations showing abnormal alterations in peripheral blood. Furthermore, functional assays indicated that treprostinil significantly inhibited tumor necrosis factor-α-induced apoptosis in human umbilical vein endothelial cells, enhanced cell viability, and alleviated endothelial inflammatory responses. In conclusion, we established a robust CAD diagnostic model and screened potential therapeutic drugs, offering new perspectives for the diagnosis and treatment of CAD.