ACAT: A Fast and Powerful P-value Combination Method for Rare-variant Analysis in Sequencing Studies

Abstract

Set-based analysis that jointly tests the association of variants in a group has emerged as a popular tool for analyzing rare and low-frequency variants in sequencing studies. The existing set-based tests can suffer significant power loss when only a small proportion of variants are causal, and their powers can be sensitive to the number, effect sizes and effect directions of the causal variants and the choices of weights. Here we propose an Aggregated Cauchy Association Test (ACAT), a general, powerful and computationally efficient p-value combination method to boost power in sequencing studies. First, by combining variant-level p-values, we use ACAT to construct a set-based test (ACAT-V) that is particularly powerful in the presence of only a small number of casual variants in a variant set. Second, by combining different variant set-level p-values, we use ACAT to construct an omnibus test (ACAT-O) that combines the strength of multiple complimentary set-based tests including the burden test, Sequence Kernel Association Test (SKAT) and ACAT-V. Through analysis of extensively simulated data and the whole-genome sequencing data from the Atherosclerosis Risk in Communities (ARIC) study, we demonstrate that ACAT-V complements the SKAT and burden test, and that ACAT-O has a substantially more robust and higher power than the alternative tests.

Convex combination sequence kernel association test for rare‐variant studies

“…We perform simulations to evaluate Type I error and compare power of our proposed test (cSKAT) and four versions of SKAT: unweighted linear combination SKAT (i.e., a sum of SKAT statistics computed separately with one annotation; Wu et al, 2013), SKAT with ideal weights equal to SNP effect sizes, and SKAT (Wu et al, 2011) and SKAT‐O (Lee et al, 2012) with weights as a function of MAF only. We also evaluate the power of a Cauchy combination test or ACAT (Liu et al, 2019) that is a combination of p values from SKAT tests for each annotation, separately. MK‐SKAT (Urrutia et al, 2016; Wu et al, 2013) software has yet to be released and, to our knowledge, is not computationally feasible for these simulations.…”

“…A wide range of SNV‐set tests have been proposed for rare‐variant association studies. Broadly speaking, they can be categorized as methods that combine SNVs (Li & Leal, 2008; Madsen & Browning, 2009; Morgenthaler & Thilly, 2007) or methods that combine marginal test statistics (or p values) for each SNV (Barnett, Mukherjee, & Lin, 2017; Conneely & Boehnke, 2007; Liu et al, 2019; Sun, Hui, Bader, Lin, & Kraft, 2019; Wu et al, 2011; Zhan et al, 2017). Methods that combine SNVs, or burden tests, evaluate the association between a trait and a weighted sum of SNVs (or burden score).…”

A gene-level methylome-wide association analysis identifies novel Alzheimer’s disease genes

“…ACAT [29,30] is a recently proposed Cauchy combination test to combine p-values. Because ACAT is an important building block for our proposed method CMO, we briefly review it here.…”

Dynamic Scan Procedure for Detecting Rare-Variant Association Regions in Whole-Genome Sequencing Studies

“…However, the true disease model is unknown and variable in practice. To robustly aggregate information from different tests and weights, we propose an omnibus test that uses the Cauchy method via ACAT, 28 and we define the test statistic as…”

“…We include three tests in the SCANG framework: the burden test (SCANG-B), SKAT (SCANG-S), and an efficient omnibus test for aggregating, via the aggregated Cauchy association test (ACAT) method, the information from the burden test and SKAT at different choices of weights (SCANG-O). 28,29 All the set-based tests share the same reduced model under the null hypothesis, and hence, the fitted null model is the same and only needs to be fit once when scanning the genome. Therefore, the computation of SCANG is highly efficient.…”