Description

GxE is intended to be a suite of R functions for implementing gene-environment interaction (GxE) tests on a genome-wide scale.

Currently, it has one function JMA2df that implements a joint 2-df SNP and SNPxE association test using GWAS summary statistics. It is based on the statistical method proposed in Manning et al (2011), and as used in Zhang et al (2021). Please refer/cite both articles if this function is used:

1. Manning, A.K., LaValley, M., Liu, C.T., ... , Dupuis, J. (2011) "Meta-analysis of gene-environment interaction: joint estimation of SNP and SNP x environment regression coefficients". Genet Epidemiol 35(1):11-18, https://doi.org/10.1002/gepi.20546

2. Zhang, W., Venkataraghavan, S., Hetmanski, J.B., ... , Ray, D., Beaty, T.H. (2021) "Detecting gene-environment interaction for maternal exposures using case-parent trios ascertained through a case with nonsyndromic orofacial cleft". Front Cell Dev Biol 9:621018, https://doi.org/10.3389/fcell.2021.621018

Key Words: Gene-environment test; GWAS summary statistics; Interaction test; Meta-analysis

Requirements

R (>= 3.0.1)

How to Install within R

require(devtools)
source_url("https://github.com/RayDebashree/GxE/blob/main/joint_metaanalysis_2df_GxE.R?raw=TRUE")

It is recommended to download/copy the stand-alone R program in this repository, save it in your local directory of choice and source() it from your local directory. When a new version of the software is available, older versions may be removed from this repository, and the above devtools::source_url() technique may not work.

Usage

Simple example

JMA2df(SNP.coef, GXE.coef, SNP.se, GXE.se, SNP.GXE.cov=NULL, trio.package.input=FALSE)

Arguments

Input	Description
SNP.coef	The vector of estimated SNP effects of size k from a 2 degree of freedom (df) gene-environment model, where k(>1) is the number of studies from which the SNP effects are obtained. In other words, a model with main effect G and interaction effect GxE are fit for each of the k independent studies to be meta-analyzed, and SNP.coef contains all the coefficient estimates for G.
GXE.coef	The vector of estimated GxE effects, from the afore-mentioned model, of size k .
SNP.se	The vector of estimated standard errors of SNP effects, from the afore-mentioned model, of size k.
GXE.se	The vector of estimated standard errors of GxE effects, from the afore-mentioned model, of size k.
SNP.GXE.cov	The vector of estimated covariance between SNP and GxE effects, from the afore-mentioned model, of size k. Default value is NULL.
trio.package.input	Logical; default value is FALSE. If TRUE, the SNP and GxE effect estimates are obtained from the R trio package (as used in Zhang et al, 2021), and no input should be provided for SNP.GXE.cov.

Value

Output	Description
SNP.coef.JMA	Jointly meta-analyzed SNP effect for a 2-df model with main effect G and gene-environment interaction effect GxE.
GXE.coef.JMA	Jointly meta-analyzed GxE effect for the afore-mentioned 2-df model.
SNP.se.JMA	Jointly meta-analyzed standard error of SNP effect estimate for the afore-mentioned 2-df model.
GXE.se.JMA	Jointly meta-analyzed standard error of GxE effect estimate for the afore-mentioned 2-df model.
SNP.GXE.cov.JMA	Jointly meta-analyzed covariance between SNP and GxE effect estimates for the afore-mentioned 2-df model.
wald2df.stat.JMA	2-df Wald test statistic for the joint test of SNP and GxE effects.
wald2df.pval.JMA	P-value of the 2-df Wald test. P-value below chosen threshold (usually the traditional genome-wide threshold of 5e-8) means either or both the SNP and the GxE effects are statistically significantly different from 0.