Comprehensive Analysis of Autophagy-Related Gene Profiles and Immune Characteristics in Parkinson's Disease.

Parkinson's disease (PD) is a progressive neurodegenerative disorder characterized by dopaminergic neuronal loss and α-synuclein pathology, yet the role of autophagy in PD and its translational relevance remain incompletely defined. Here, we integrated GEO transcriptomic data sets (GSE7621, GSE20141, and GSE20163) to identify autophagy-related biomarkers and to delineate their biological context. Differentially expressed genes (DEGs) were intersected with autophagy genes curated in the Human Autophagy Database to derive autophagy-related DEGs, followed by Gene Ontology and KEGG analyses. Biomarkers were prioritized using three complementary machine-learning approaches and validated in an independent cohort (GSE49036), with a nomogram constructed for PD risk estimation. We further profiled biomarker-associated programs using gene set enrichment analysis, ceRNA network inference, and immune-cell infiltration assessment, and evaluated expression changes in an MPTP-induced PD mouse model. We identified 2177 DEGs across discovery data sets, and intersection with 232 autophagy genes yielded 29 autophagy-related DEGs. Machine-learning analyses nominated six hub genes (PEX14, VEGFA, BECN1, LAMP1, CXCR4, and ATF6), among which external validation robustly supported PEX14 and LAMP1. Enrichment analyses linked these markers to immune and inflammation pathways, including cytokine-cytokine receptor interaction, antigen processing and presentation, and hematopoietic cell lineage, and suggested concomitant shifts in immune infiltration. Consistently, MPTP-treated mice exhibited decreased PEX14 and increased LAMP1 expression. Together, our findings identify PEX14 and LAMP1 as autophagy-related biomarkers in PD, connecting autophagy with immune-related signaling and supporting their potential utility for biomarker-based PD risk prediction and therapeutic stratification.