Differential dropout among SNP genotypes and impacts on association tests

Area of Science:

• Genetics
• Biotechnology
• Statistical Genomics

Background:

• High-throughput genotyping technologies achieve high accuracy and call rates.
• Concerns exist regarding genotype-dependent performance biases in genetic association studies.
• Differential dropout rate is defined as the ratio of no-call rates between heterozygotes and homozygotes.

Purpose of the Study:

• To examine the impact of differential genotype dropout on association tests.
• To evaluate methods for detecting differential dropout.
• To assess the magnitude of differential dropout in public datasets.

Main Methods:

• Simulation studies of population- and family-based association tests.
• Investigation of Hardy-Weinberg Equilibrium (HWE) and call rate-heterozygosity correlation for detection.
• Analysis of two public genotype datasets to quantify differential dropout.

Main Results:

• Differential dropout has minimal impact on case-control association tests.
• Family-based tests can experience severe bias, particularly with low disease allele frequencies (e.g., 5%).
• A differential dropout rate of 2.5 significantly biased results even at 98% call rate; HapMap data showed detectable differential dropout.

Conclusions:

• Case-control association studies are robust to differential dropout.
• Family-based association tests are highly susceptible to bias from differential dropout.
• Researchers should actively control for differential dropout, even with high-accuracy, high-call-rate genotype data, as seen in HapMap data.