Applying AlphaGenome to Autoimmune Disease: Predicting Lupus-Associated Gene Expression Across Immune Cell Types

Abstract

Systemic lupus erythematosus (SLE) is a chronic autoimmune disease where the immune system mistakenly attacks healthy cells and tissues, causing inflammation and organ damage throughout the body. While over 300 different genetic risk loci have been identified through genome-wide association studies (GWAS), cell-type specific patterns of gene expression remain  poorly understood. Our work provides a case study on how to identify distinct gene expression signatures across different immune cell types and associated with different genetic variants by using AlphaGenome, an AI tool developed by Google Deepmind. Three lupus associated genes — TNFRSF17, STAT4, and IRF5 — were analyzed across B cells, T cells, and kidney tissues. Our results show that a lupus-associated SNP in TNFRSF17 increases gene expression in B cells, that STAT4 is expressed across kidney, T cells, and B cells, and that IRF5 shows similar high expression in both B and T cells, remaining consistent with their roles in lupus and immune signaling. These findings demonstrate how multiple lupus risk genes affect the same immune cell types, each gene contributing to its own layer of immune dysfunction.