Related Experiment Videos
Population structure, differential bias and genomic control in a large-scale, case-control association study
David G Clayton1, Neil M Walker, Deborah J Smyth
1Juvenile Diabetes Research Foundation/Wellcome Trust Diabetes and Inflammation Laboratory, University of Cambridge, Cambridge Institute for Medical Research, Wellcome Trust/MRC Building, Cambridge, CB2 2XY, UK.
Nature Genetics
|October 18, 2005
Summary
Population structure and laboratory bias inflated genetic association test statistics in type 1 diabetes research. A modified genomic control method was developed to address these issues, improving the reliability of genetic association studies.
Area of Science:
- Genetics
- Epidemiology
- Bioinformatics
Background:
- Causal inference in epidemiological case-control studies faces challenges like confounding, selection bias, and misclassification.
- Population structure is a significant confounding factor in genetic association studies.
- Previous studies have shown inflation in test statistics due to these factors.
Purpose of the Study:
- To identify and quantify sources of inflation in test statistics for genetic association studies.
- To propose a method to mitigate the impact of confounding and bias on genetic association findings.
- To improve the accuracy of causal inference in genetic epidemiology.
Main Methods:
- Analysis of 6,322 nonsynonymous SNPs in 816 type 1 diabetes cases and 877 controls.
- Quantification of test statistic inflation attributed to population structure.
- Identification of differential genotype scoring bias from laboratory differences.
- Extension of the genomic control method with variable SNP downweighting.
Main Results:
- A +11.2% inflation in test statistics was observed.
- Population structure accounted for a portion of this inflation.
- Differential laboratory bias in genotype scoring contributed to false-positive associations.
- The extended genomic control method effectively addressed SNP downweighting.
Conclusions:
- Both population structure and laboratory-induced bias can significantly inflate genetic association test statistics.
- The extended genomic control method provides a robust approach to manage these biases.
- This methodology enhances the reliability of findings in genetic association studies, particularly for complex diseases like type 1 diabetes.