Estimating missing heritability for disease from genome-wide association studies

Area of Science:

• Genetics
• Statistical genetics
• Complex trait analysis

Background:

• Genome-wide association studies (GWAS) aim to find single nucleotide polymorphisms (SNPs) associated with complex traits.
• Traditional methods using strict significance thresholds for individual SNPs can lead to false negatives.
• A prior method was developed to estimate variation accounted for by all SNPs simultaneously in quantitative traits.

Purpose of the Study:

• To extend a simultaneous SNP analysis method for case-control studies.
• To accurately estimate the heritability explained by common SNPs for binary traits.
• To apply the developed method to real-world case-control data.

Main Methods:

• Utilized a linear mixed model for binary trait analysis.
• Transformed estimates to a liability scale, adjusting for trait scale and case ascertainment.
• Validated the method's unbiasedness through theoretical analysis and simulations.

Main Results:

• The developed method provides unbiased estimates of variation accounted for by common SNPs.
• Applied to Wellcome Trust Case Control Consortium data, the method revealed significant SNP-tagged liability variation.
• Substantial proportions of liability variation for Crohn disease, bipolar disorder, and type I diabetes were identified.

Conclusions:

• The enhanced method is effective for estimating SNP-heritability in case-control studies.
• Common SNPs tag a significant portion of liability variation for major complex diseases.
• This approach improves the understanding of genetic architecture in complex diseases.