Abstract:
|
Individuals with pedigree or population structure are usually included in genetic association studies. Due to linkage disequilibrium and familial relationship, genotype data from such studies often have complex correlations, which need to be taken into account appropriately. The Adaptive Burden Test (ABT) has been developed for multi-variant association testing in this situation. While this method shows advantages in boosting power via "data-driven" weights, its unique feature in handling missingness in genotype data has not been explored. In this work we demonstrate that ABT is able to impute missing genotype using the best linear unbiased predictor (BLUP). By taking advantage of the dependence among markers and pre-known relatedness among individuals, missing genotype can be imputed, which provides additional power for detecting association. We demonstrate the performance of the built-in BLUP imputation of ABT in a simulation study, and apply this approach to analyze the fasting glucose data from the Framingham Heart study.
|