Multi-omics profiling uncovers candidate genes for the treatment of graves' disease.

BACKGROUND

This Summary-data-based Mendelian randomization (SMR) study investigates the genetic basis of Graves' disease (GD) using a multi-omics approach to elucidate biological mechanisms and identify therapeutic candidates.

METHODS

SMR and Bayesian colocalization analyzed GD associations with methylation (mQTL), expression (eQTL), and protein quantitative trait loci (pQTL) using FinnGen as the discovery cohort. Top SNPs underwent methylation-gene expression and gene expression-protein interaction analyses. Key findings were validated against an independent GWAS dataset. Functional enrichment characterized pathways for eQTL-SMR-derived loci. The Drug Signatures Database (DsigDB) predicted potential drugs.

RESULTS

Significant multi-omics associations were found. mQTL analysis linked GD to 1,772 methylation sites (294 genes); eQTL identified 157 genes. pQTL signals appeared in two cohorts (4 genes deCODE; 16 genes UKB-PPP). AGER and AIF1 were consistent across omics. AIF1 showed strong GD associations (mQTL OR 1.09; eQTL OR 1.44; deCODE-pQTL OR 3.17; UKB-PPP-pQTL OR 3.46). Integrated analysis revealed a positive correlation between AGER DNA methylation and expression, but negative for AIF1, with colocalization support (PPH3 + PPH4 > 0.8). SMR showed inverse pQTL associations for AGER and positive for AIF1, also colocalized. Independent GWAS validation confirmed GD associations with AGER and AIF1 at pQTL. Enrichment indicated immune/inflammatory pathways. Drug predictions included agents for AGER (Oleanolic Acid, Rutin, Estriol, GW9662) and AIF1 (Quinoline).

CONCLUSION

This study provides novel insights into GD pathogenesis, highlighting AGER and AIF1 as therapeutically relevant genes. Findings offer valuable insights for GD treatment strategies and potential drug development streamlining, warranting further experimental validation.