Machine Learning Integrates Bulk and Single-Nucleus RNA Sequence to Explore Apoptosis-Related Gene in Myocardial Infarction.

Abstract

BACKGROUND: Myocardial infarction (MI) remains a predominant contributor to global cardiovascular disease death, with apoptosis playing a pivotal yet incompletely characterized role in adverse cardiac remodeling. Despite advances in bulk RNA sequence analyses, the single-cell level analyses of apoptotic signaling of MI progression remain unclear. METHODS: Multiomics data integration was synthesized through single-nucleus RNA-seq (snRNA-seq) (GSE270788) and bulk RNA-seq (GSE21610). Three hundred ninety-seven ARGs (apoptosis-related genes) were downloaded from GeneCards (relevance score > 7), followed by identification of candidate genes through AUCell scoring and correlation analysis. Random Forest, Boruta, and LASSO algorithms were employed to refine hub genes. Hub genes were validated by animal and cell experiments. RESULTS: Single-nucleus sequence analysis reveals dramatic variation in ARGs activity across cardiac cells after MI, with fibroblasts demonstrating significantly elevated apoptotic activity compared with other cell types. Integration of single-nucleus and bulk sequence confirms TNFRSF12A as the major molecule regulating ARGs activation in MI. Experiment validation confirmed that upregulation of TNFRSF12A in the MI model and TGF-β induced NIH3T3. CONCLUSION: This study revealed that fibroblast dominates apoptotic activity post-MI. We identified TNFRSF12A as a key regulator in MI through multiomics analysis, with potential as a diagnostic biomarker and therapeutic target.