Abstract: TH-OR031
Atlas of Glomerular Disease-Specific Genetic Effects on Gene Regulation in Blood Empowers New Gene Discovery Studies
Session Information
- Genetics of Complex Kidney Traits
November 06, 2025 | Location: Room 360A, Convention Center
Abstract Time: 05:00 PM - 05:10 PM
Category: Genetic Diseases of the Kidneys
- 1202 Genetic Diseases of the Kidneys: Complex Kidney Traits
Authors
- Liu, Lili, Columbia University, New York, New York, United States
- Wang, Chen, Columbia University, New York, New York, United States
- Eichinger, Felix H., University of Michigan, Ann Arbor, Michigan, United States
- Fermin, Damian, University of Michigan, Ann Arbor, Michigan, United States
- Sanna-Cherchi, Simone, Columbia University, New York, New York, United States
- Sampson, Matt G., Boston Children's Hospital, Boston, Massachusetts, United States
- Gbadegesin, Rasheed A., Duke University, Durham, North Carolina, United States
- Troyanskaya, Olga, Princeton University, Princeton, New Jersey, United States
- Gharavi, Ali G., Columbia University, New York, New York, United States
- Kretzler, Matthias, University of Michigan, Ann Arbor, Michigan, United States
- Kiryluk, Krzysztof, Columbia University, New York, New York, United States
Background
IgA nephropathy (IgAN), focal segmental glomerulosclerosis (FSGS), membranous nephropathy (MN), and minimal change disease (MCD) account for the majority of idiopathic glomerulopathies (GN). However, there are no powered transcriptomic datasets coupled to genetic data to investigate the genetic mechanisms underlying gene regulation in the context of GN.
Methods
We performed genome sequencing and blood RNA-seq on 1,826 participants from the CureGN study, a prospective cohort of primary GN. We generated transcriptome-wide maps of eQTL, sQTL, and edQTL effects for FSGS (N=450), IgAN (N=403), IgA vasculitis (N=123), MCD (N=408), and MN (N=442). QTL and interaction QTL mapping were performed using tensorQTL, adjusting for age, sex, genetic PCs, and PEER factors, with and without adjustment for deconvoluted cell fractions. We constructed cross-disease QTL maps (N=1826) adjusted for GN type. Leveraging context-specific eQTLs we built GN type-specific predictive models of gene expression for disease-specific TWAS.
Results
Context-specific QTL mapping identified 13,117 eGenes, 4,644 sGenes and 56 edGenes with an FDR<0.05 in at least one GN type. Approximately 5%-10% of the QTLs were unique to a specific GN type, and ~90% were shared between at least two conditions. Colocalization analysis suggested that ~80% of shared eGenes between traits also shared the same casual variants (PP4>=80%), whereas ~2% had distinct causal variants (PP3>=80%) indicating context-specific regulatory effects. Context-specific TWAS identified novel candidate genes significantly associated with GN risk. Cross-phenotype QTL mapping uncovered 17,210 eGenes and 5,145 sGenes, including 6,466 eGenes and 2,705 sGenes not previously detected in GTEx. Cell-type-specific interaction QTL mapping refined a set of QTL signals specific to each cell type and GN form. Interaction QTL analyses for age, sex, eGFR, and proteinuria identified QTL effects that are age, sex, and disease severity specific.
Conclusion
Using the CureGN dataset, we generated comprehensive maps of cell-type specific and GN-context-specific genetic effects on gene expression, splicing and double-strand RNA editing, providing a powerful resource for integrative gene discovery studies of primary GN.