Comprehensive analysis of autophagy status and its relationship with immunity and inflammation in ischemic stroke through integrated transcriptomic and single-cell sequencing.

Ischemic stroke (IS) significantly impacts patients' health and quality of life, with the roles of autophagy and autophagy-related genes in IS still not fully understood. In this study, IS datasets were retrieved from the GEO database. Autophagy-related genes(ARGs) were identified and screened for differential expression. A prediction model was constructed using machine learning algorithm. WGCNA was employed to analyze differential regulation modules among different clusters of stroke patients. The analysis results were validated using single-cell sequencing data. Finally, autophagy hub genes were validated in an external cohort and an IS mouse model. We observed suppressed autophagy states in IS patients. A diagnostic model with good clinical efficacy for stroke diagnosis was constructed based on the selected key genes (AUC = 0.87). Consensus clustering identified two IS subtypes with distinct gene expression patterns and immune cell infiltration. scRNA-seq data analysis confirmed downregulation of pexophagy in IS. CellChat analysis identified key signaling pathways and intercellular interactions related to pexophagy. Validation in an external cohort and IS mouse model confirmed differential gene expression, supporting the involvement of pexophagy in IS pathogenesis. The identified key genes, molecular subtypes, and cellular interactions provide a foundation for further research into targeted therapies and precision medicine approaches for IS patients.