Gene-based association tests using GWAS summary statistics

Abstract

Motivation A huge number of genome-wide association studies (GWAS) summary statistics freely available in databases provide a new material for gene-based association analysis aimed at identifying rare genetic variants. Only a few of the many popular gene-based methods developed for individual genotype and phenotype data are adapted for the practical use of the GWAS summary statistics as input. Results We analytically prove and numerically illustrate that all popular powerful methods developed for gene-based association analysis of individual phenotype and genotype data can be modified to utilize GWAS summary statistics. We have modified and implemented all of the popular methods, including burden and kernel machine-based tests, multiple and functional linear regression, principal components analysis and others, in the R package sumFREGAT. Using real summary statistics for coronary artery disease, we show that the new package is able to detect genes not found by the existing packages. Availability and implementation The R package sumFREGAT is freely and publicly available at: https://CRAN.R-project.org/package=sumFREGAT. Supplementary information Supplementary data are available at Bioinformatics online.

Prioritization of causal genes for coronary artery disease based on cumulative evidence from experimental and in silico studies

“…We also extracted data on the prioritized genes from four previously published in silico studies [46][47][48][49] . A brief summary of approaches used in that studies is provided in Supplementary Methods.…”

Gene-based mapping of trehalose biosynthetic pathway genes reveals association with source- and sink-related yield traits in a spring wheat panel

“…We therefore recommend using both analyses to maximize the overall number of variants/genes detected. We also observed that the multiple linear regression approach often outperformed the methods with random-effects (Figure 2a) showing relatively good model fit (see QQ plots in Figure S1) although false positives are often reported for fixed-effect models (Deng and Pan, 2018;Svishcheva et al, 2019).…”

“…The weights were calculated using the standard probability density function of the beta distribution. For further detail on the model description see Svishcheva et al (2019). We estimated the P values by using Kuonen's method (Kuonen, 1999) and considered the mode of inheritance as additive.…”

<i>In silico</i> mapping of coronary artery disease genes

“…Подробно методы и их имплементация описаны в на шей работе (Svishcheva et al, 2019), где также проведена апробация каждого метода на выборках различной структуры с использованием реальных данных, в том числе MICAD и UKbb.…”

“…Такие данные никогда не депонируются в открытые базы данных, в отличие от результатов ПГАА и метаанализа, представ ленных набором суммарных статистик для каждого исследованного генетического варианта. В последних работах было показано, что РАА можно проводить in silico, используя суммарные статистики, не требуя индивидуальных фенотипов и генотипов (de Leeuw et al, 2015;Bakshi et al, 2016;Wang et al, 2017;Svishcheva et al, 2019).…”