Detecting Rare Variants in Case-Parents Association Studies

Abstract

Despite the success of genome-wide association studies (GWASs) in detecting common variants (minor allele frequency ≥0.05) many suggested that rare variants also contribute to the genetic architecture of diseases. Recently, researchers demonstrated that rare variants can show a strong stratification which may not be corrected by using existing methods. In this paper, we focus on a case-parents study and consider methods for testing group-wise association between multiple rare (and common) variants in a gene region and a disease. All tests depend on the numbers of transmitted mutant alleles from parents to their diseased children across variants and hence they are robust to the effect of population stratification. We use extensive simulation studies to compare the performance of four competing tests: the largest single-variant transmission disequilibrium test (TDT), multivariable test, combined TDT, and a likelihood ratio test based on a random-effects model. We find that the likelihood ratio test is most powerful in a wide range of settings and there is no negative impact to its power performance when common variants are also included in the analysis. If deleterious and protective variants are simultaneously analyzed, the likelihood ratio test was generally insensitive to the effect directionality, unless the effects are extremely inconsistent in one direction.

Explaining additional genetic variation in complex traits

“…Even when considering transmission within-families, which are likely to complement association studies, large numbers of families will be required [72]. Second, these variants are less likely to be in strong LD with common variants that are tagged on current SNP chips because (i) they may be under stronger selection and therefore be younger polymorphisms with lower minor allele frequency; and (ii) many may be deletions or duplications (i.e.…”

Adaptive Combination of P-Values for Family-Based Association Testing with Sequence Data

“…Many statistical methods were proposed for rare variant association testing, but most of them were designed for population-based studies. Among the family-based rare variant association testing methods, some extend the transmission disequilibrium test [56] , [57] and focus on parent-child trio data [36] , [37] . Some methods are eligible for analyzing pedigree data (including but not limited to trios), and they can be categorized as the burden tests and the non-burden tests (e.g., famSKAT [31] , [38] , FFBSKAT [39] , Kernel [32] ).…”