Musani, S. | Search bioRxiv

bioRxiv2019

Sequencing and Imputation in GWAS: Cost-Effective Strategies to Increase Power and Genomic Coverage Across Diverse Populations

A key aim for current genome-wide association studies (GWAS) is to interrogate the full spectrum of genetic variation underlying human traits, including rare variants, across populations. Deep whole-genome sequencing is the gold standard to capture the full spectrum of genetic variation, but remains prohibitively expensive for large samples. Array genotyping interrogates a sparser set of variants, which can be used as a scaffold for genotype imputation to capture variation across a wider set of variants. However, imputation coverage and accuracy depend crucially on the reference panel size and genetic distance from the target population. Here, we consider a strategy in which a subset of study participants is sequenced and the rest array-genotyped and imputed using a reference panel that comprises the sequenced study participants and individuals from an external reference panel. We systematically assess how imputation quality and statistical power for association depend on the number of individuals sequenced and included in the reference panel for two admixed populations (African and Latino Americans) and two European population isolates (Sardinians and Finns). We develop a framework to identify powerful and cost-effective GWAS designs in these populations given current sequencing and array genotyping costs. For populations that are well-represented in current reference panels, we find that array genotyping alone is cost-effective and well-powered to detect both common- and rare-variant associations. For poorly represented populations, we find that sequencing a subset of study participants to improve imputation is often more cost-effective than array genotyping alone, and can substantially increase genomic coverage and power.

genetics↗

bioRxiv2018

A Fully-Adjusted Two-Stage Procedure for Rank Normalization in Genetic Association Studies

When testing genotype-phenotype associations using linear regression, departure of the trait distribution from normality can impact both Type I error rate control and statistical power, with worse consequences for rarer variants. While it has been shown that applying a rank-normalization transformation to trait values before testing may improve these statistical properties, the factor driving them is not the trait distribution itself, but its residual distribution after regression on both covariates and genotype. Because genotype is expected to have a small effect (if any) investigators now routinely use a two-stage method, in which they first regress the trait on covariates, obtain residuals, rank-normalize them, and then secondly use the rank-normalized residuals in association analysis with the genotypes. Potential confounding signals are assumed to be removed at the first stage, so in practice no further adjustment is done in the second stage. Here, we show that this widely-used approach can lead to tests with undesirable statistical properties, due to both a combination of a mis-specified mean-variance relationship, and remaining covariate associations between the rank-normalized residuals and genotypes. We demonstrate these properties theoretically, and also in applications to genome-wide and whole-genome sequencing association studies. We further propose and evaluate an alternative fully-adjusted two-stage approach that adjusts for covariates both when residuals are obtained, and in the subsequent association test. This method can reduce excess Type I errors and improve statistical power.

genetics↗