Identification of Biomarkers for Sjögren’s Syndrome via Bioinformatics Analysis and Mendelian Randomization

The data used in this study was sourced from three independent peripheral blood transcriptome datasets of Sjogren's syndrome (SS) in the GEO database: GSE51092 (190 SS patients, 32 controls), GSE66795 (131 SS patients, 29 controls), and GSE84844 (30 SS patients, 30 controls). All samples were collected from human peripheral blood between 2013 and 2017, involving multiple research centers in the United States, United Kingdom, and Japan. The data processing was completed using R software (version 4.4.1). The original expression data was normalized using RMA method and batch effects were corrected using ComBat algorithm in the "sva" software package. Principal Component Analysis (PCA) is used to verify the effectiveness of batch calibration. The integrated dataset contains a total of 442 samples (351 SS patients and 91 controls), and 54 differentially expressed genes were identified through differential expression analysis (| log2FC | ≥ 0.585, adjusted p<0.05). There are no missing data in the processed expression matrix. The main data files include: (1) a list of 26152 SNPs for the instrumental variables used in Mendelian randomization analysis; (2) Supplementary Table S2 contains complete MR analysis results of 195 SS related genes; (3) Scatter plots, forest plots, and sensitivity analysis plots of 5 candidate genes, and Supplementary Figure S1. All processed data files can be opened using standard software, such as Microsoft Excel to open table files and Adobe Acrobat Reader to open image files. R scripts for data processing and analysis can be provided upon request.