Karl Broman (following the code in Nick Furlotte's pylmm)
In order to better understand pylmm, for linear mixed models for genome-wide association studies (GWAS), I wrote an R version of the algorithm.
-
Results compared to pylmm (and also the R package regress). [source]
-
Performance compared to pylmm. [source]
You can install R/lmmlite from GitHub.
You first need to install the remotes pakcage.
install.packages("remotes")
Then use install_github() to install R/lmmlite.
library(remotes)
install_github("kbroman/lmmlite")
The Rcpp and RcppEigen packages will also be installed.
R/lmmlite is released under the GNU Affero GPL.
The code was developed following study of Nick Furlotte's pylmm code.