Abstract:
|
Identifying potentially pathogenic rare variants requires a large sample size. Using external controls, whose genomes have been sequenced and publicly available, could be a cost effective approach to increase the power of rare variant testing. The integrating External Controls into Association Testing (iECAT) method uses an empirical Bayesian approach and improves statistical power while controlling for type I error rates attributable to batch effects from incorporating disparate external study data. However, iECAT cannot adjust for covariates or exploit batch effect information presented in across SNPS in its originally presented form. We develop a new score test based on the Empirical Bayesian framework, which can adjust for covariates. Specifically, we compare the batch effect between internal and external control samples, and construct a statistic using weighted statistics of internal and combined samples. We then extend our method to incorporate batch effect information across SNPs and QC annotations. We show by simulation studies that our method has increased power over the original iECAT across a wide range of scenarios while controlling for type I error rates.
|