Impact of fine-scale UK population structure on genome-wide association studies of complex human traits
Principal Investigator:
Professor Andrew Morris
Approved Research ID:
15390
Approval date:
December 1st 2015
Lay summary
The United Kingdom (UK) is a genetically diverse nation with strong genetic differences between regions, which may adversely affect genome-wide association studies of complex human traits if not accounted for in the analysis. The aim of the proposed project is to evaluate the impact of population structure on genetic association studies of complex traits with common and rare variants, and to compare approaches to account for ancestral confounding. Our study will inform the application of methodology and interpretation of results of genome-wide association studies of complex human traits using UK Biobank data, thereby minimising the potential for false positive findings and increasing power to detect true association signals. In this way, our study will be relevant to genetic studies of all clinical outcomes considered in UK Biobank, with the potential to lead to the prevention, diagnosis and treatment of disease. We will consider a range of methodologies to detect population structure (based on participant postcode data), and assess their suitability for correction for ancestral confounding. We will identify genetic variants that are confounded with ancestry, and assess their impact on association analyses with physical measures (blood pressure, anthropometry and spirometry) taken at the assessment centre, and self-reported type 2 diabetes, myocardial infraction and asthma from the verbal interview. We will also use the genotype data to simulate traits under alternative models of structure, and assess the impact on association findings for both common and rare variants. This analysis will include the full cohort.