Diagnostic markers and potential therapeutic agents for Sjögren's syndrome screened through multiple machine learning and molecular docking.

This article focuses on identifying diagnostic biomarkers and potential therapeutic agents for primary Sjögren's syndrome (pSS), a chronic autoimmune disease affecting exocrine glands, using machine learning and molecular docking approaches. By analyzing gene expression datasets from the Gene Expression Omnibus, the study screened 1,643 differentially expressed genes and applied four machine learning methods—support vector machine (SVM), least absolute shrinkage and selection operator (LASSO) regression, random forest, and weighted correlation network analysis (WGCNA)—to identify 10 hub genes with strong diagnostic potential for pSS. Immunohistochemistry on labial gland tissues from patients validated the expression patterns of key hub genes such as EZH2, ELAVL1, and IGF1R. Additionally, immune cell infiltration analysis revealed correlations between these hub genes and specific immune cell populations, while molecular docking identified small-molecule compounds with promising binding affinities to hub gene products, suggesting new avenues for targeted therapy in pSS. The study provides a foundation for improved early diagnosis and personalized treatment strategies but notes that further experimental validation is required.