Examining the rare disease assumption used to justify HWE testing with control samples

Area of Science:

• Population Genetics
• Statistical Genetics
• Genomic Association Studies

Background:

• Genome-wide association studies (GWAS) and other genetic analyses often assume Hardy-Weinberg Equilibrium (HWE).
• HWE testing is crucial for validating statistical assumptions in genetic data analysis.
• In case-control studies, HWE can be distorted if a genetic marker is associated with the disease, leading to HWE testing on controls only.

Purpose of the Study:

• To investigate the validity of testing Hardy-Weinberg Equilibrium (HWE) using only control samples in genetic studies.
• To address the implications of the rare disease assumption and potential issues with control-only HWE testing across different disease prevalences.

Main Methods:

• Theoretical derivations were performed to analyze the impact of control-only HWE testing.
• Numerical studies utilized simulated genotypes and real data from the 1000 Genomes Project.
• The study evaluated the Type I error rate under various conditions.

Main Results:

• Testing for Hardy-Weinberg Equilibrium (HWE) using only control samples leads to severely inflated Type I error rates.
• This inflation occurs irrespective of whether the disease under investigation is rare or common.
• The findings indicate that the common practice of control-only HWE testing is statistically unsound.

Conclusions:

• The reliance on control-only HWE testing can result in the erroneous discarding of numerous genetic variants.
• This practice wastes valuable genetic information and impedes the identification of variants associated with diseases.
• Re-evaluation of HWE testing strategies in genetic association studies is warranted.