Approved research

Genome-wide association studies of kidney diseases and phenome-wide association studies of putative targets

Lay summary

Chronic kidney disease (CKD) affects 10% of the world's population and can lead to dialysis or transplant. Despite the urgent need for targeted therapeutics, the understanding of the genetic basis of CKD has lagged other diseases (e.g. cancer) for decades. To identify genetic causes of kidney diseases, we have built a database of clinical and genetic data from 3000 patients and >20000 disease-free controls. We plan to use the UK Biobank data for two main aims: (1) to replicate our findings in an independent dataset of additional individuals with kidney disease and (2) to infer safety profiles for drug targets by examining traits of 'human knockouts' - individuals with potentially damaging mutations that may lack severe disease. The duration of the project will be over 3 years, with annual renewals, to enable ongoing assessment of genetic predisposition to disease and safety profiles. For aim (1) we will conduct two main analyses: a GWAS of common variants using the genotype chip data and a rare variant aggregation test using the WES data. The common variant GWAS will be a standard logistic regression at each common variant passing quality control, including all patients with diabetic nephropathy as cases and all T2D patients without nephropathies as controls, and controlling for ancestry, sex, age, duration of diabetes, hypertension, and any other clinical factors known to contribute to nephropathy. Rare variants will be aggregated within genes and tested for associations with disease status using burden and SKAT tests, controlling for same confounders as above. For aim (2) we will conduct a PheWAS for a putative target gene with logistic regression, comparing all loss-of-function (LoF) variant carriers in the entire dataset to those without any LoF variants. We will test all phenotypes (e.g. ICD 10 codes) for association with LoF variant carrier status. For all analyses, we will control inflation of the Type I error rate using Bonferroni correction for multiple testing.