Genetic, Transcriptomic, and Epigenomic Insights into Sjögren's Disease: An Integrative Network Investigation and Immune Diseases Comparison

Abstract

Sjögren's disease (SjD) is a systemic autoimmune disorder primarily causing dry eyes and mouth. It frequently overlaps with other autoimmune diseases (AIDs), including rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE). However, the genetic basis of SjD remains underexplored, limiting our understanding of its connections to other immune-mediated conditions. In this study, we aimed to identify gene networks associated with SjD through the integration of genetic, transcriptomic, and epigenomic data. We further compared the genetic factors of SjD with other immune-mediated diseases. We analyzed genome-wide association studies (GWAS) summary statistics, DNA methylation, and transcriptomic data using our in-house network-based methods, dmGWAS and EW_dmGWAS, to identify key gene modules associated with SjD. In dmGWAS analysis, discovery and evaluation datasets were used to identify consensus results. We conducted gene-set, cell-type, and disease-enrichment analyses on significant gene modules and explored potential drug targets. Genetic correlations and Mendelian randomization were applied to assess SjD's link with 17 other AIDs and 16 cancer types. dmGWAS identified 207 and 211 gene modules in the discovery and evaluation phases, respectively, while EW_dmGWAS detected 886 modules. Key modules highlighted 55 genes (discovery), 52 genes (evaluation), and 59 genes (EW_dmGWAS), with at least 50 genes from each analysis linked to AIDs and cancer. Enrichment analyses confirmed their relevance to immune and oncogenic pathways. We pinpointed four candidate drug targets associated with AIDs. We developed a novel integrative omics approach to identify potential genetic markers of SjD and compared them with AIDs and cancers. Our approach can be similarly applied to other disease studies.

Keywords: Mendelian randomization; autoimmune disease; drug target; genetic correlation; genome-wide association studies; pleiotropy.

Figure 1

Workflow of the study design. We integrated the epigenomic and transcriptomic data with the GWAS summary statistics using network-based analysis to identify the top gene networks associated with SjD. We further selected the top gene networks for downstream analysis. We also compared the genetic and causal relationship of SjD with other diseases.

Figure 2

Protein—protein interaction (PPI) subnetworks across the selected top gene modules: (A) PPI network of the dmGWAS module genes using the discovery data; (B) PPI network of the dmGWAS module genes using the evaluation data; (C) PPI network of EW_dmGWAS module genes.

Figure 3

Over—representation analysis (ORA), disease—enrichment analysis (DEA), and drug—target enrichment analysis (DTEA) on the selected top gene modules: (A) ORA of the top gene modules from the discovery data based on dmGWAS analysis; (B) ORA of the top gene modules from the evaluation data based on dmGWAS analysis; (C) ORA of the top gene modules from the transcriptomic data based on the EW_dmGWAS analysis; (D) DTEA of the top gene modules from the discovery data based on dmGWAS analysis; (E) DTEA of the top gene modules from the discovery data based on dmGWAS analysis; (F) ORA of the top gene modules from the transcriptomic data based on the EW_dmGWAS analysis; (G) DTEA of epigenomic and transcriptomic genes.

Figure 4

Heatmaps of genetic correlation analyses: (A) genetic correlation of SjD and 17 other autoimmune diseases (AIDs); (B) genetic correlation of SjD with 16 cancer types. *—denotes the diseases are statistically significant at p-value < 0.05. Highlighted red rectangle bars show the genetic correlation of SjD with 17 other AIDs.

Figure 5

Forest plot of Mendelian randomization analysis: (A) Mendelian randomization of SjD with 17 other autoimmune diseases (AIDs); (B) Mendelian randomization of SjD with 16 cancer types.