Skip to content
An R package for detecting Gene-by-Environment (GxE) interaction effects on the transcriptome using Allele Specific Expression quantified from RNA-seq
R C++ Stan
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.
R fixed printing of det==0 loci Jan 20, 2016
beta_binomial_models Added beta binomial GLM and GLMM Dec 2, 2015
man check for NAs in alt, y, xFull and xNull Nov 24, 2015
src Merge branch 'master' of Nov 24, 2015
NAMESPACE Final paper version Aug 27, 2015 Update Aug 1, 2017
big_test.R removed optimization flags for kasper :) Sep 2, 2015
test.R Added more realistic test, with filtering Aug 28, 2015
tss.txt.gz Added example script for candidate variant mapping using EAGLE Dec 2, 2015

eagle: Environment-ASE through Generalized LinEar modeling

EAGLE is an R package which uses Allele Specific Expression (ASE) quantified via RNA-seq to detect gene by environment (GxE) interactions. The underlying idea is that observing an association between the level of allelic imbalance at an exonic SNP, and an environmental factor, is evidence of a cis-regulatory element modulating the transcriptomic response to the environmental perturbation. The underlying statistical model is a binomial Generalized Linear Mixed Model, with a random effect term used to account for count overdisperion. Model fitting is achieved using Non-conjugate Variational Message Passing.

The paper describing EAGLE has now been published:

Allele-specific expression reveals interactions between genetic variation and environment. Knowles, D. A; Davis, J. R; Edgington, H.; Raj, A.; Favé, M.; Zhu, X.; Potash, J. B; Weissman, M. M; Shi, J.; Levinson, D.; Awadalla, P.; Mostafavi, S.; Montgomery, S. B; and Battle, A. Nature Methods 2017.

An early bioRxiv preprint is available.

The code is on github.

System requirements

eagle has been tested under the following architectures:

  • Mac OS X Yosemite 10.10.2, R 3.1.2
  • Ubuntu 14.04 LTS, R 3.2.2
  • Red Hat 4.4.7, R 3.1.2

I have not tried compiling under Windows.


You will need the RcppEigen R package installed. In R run


To download the code

git clone

To compile+install, in this directory run

R CMD INSTALL --build .

Alternatively run

# require(devtools)


For a simple example script on synthetic data look at test.R. For a slightly more involved/realistic example, including eQTLs and realistic filtering options look at big_test.R.

You can’t perform that action at this time.