Misspecified Models for Gene-Environment Interaction

First major functionality is to compute the bias in misspecified linear gene-environment interaction models. The most generalized function for this objective is GE_bias(). However GE_bias() requires specification of many higher order moments of covariates in the model. If users are unsure about how to calculate/estimate these higher order moments, it may be easier to use GE_bias_normal_squaredmis(). This function places many more assumptions on the covariates (most notably that they are all jointly generated from a multivariate normal distribution) and is thus able to automatically calculate many of the higher order moments automatically, necessitating only that the user specify some covariances. There are also functions to solve for the bias through simulation and non-linear equation solvers, these can be used to check your work. Second major functionality is to implement the Bootstrap Inference with Correct Sandwich (BICS) testing procedure, which we have found to provide better finite-sample performance than other inference procedures for testing GxE interaction. More details on these functions are available in Sun, Carroll, Christiani, and Lin, "Testing for Gene-Environment Interaction Under Exposure Misspecification" (Submitted).


News

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.

install.packages("GEint")

0.1.2 by Ryan Sun, 3 months ago


Browse source code at https://github.com/cran/GEint


Authors: Ryan Sun [aut, cre]


Documentation:   PDF Manual  


GPL-3 license


Imports mvtnorm, nleqslv, pracma, speedglm, rje, geepack, stats

Suggests knitr, rmarkdown


See at CRAN