No description, website, or topics provided.
Switch branches/tags
Nothing to show
Clone or download
Latest commit 21a9902 Apr 2, 2015
Failed to load latest commit information.
warpedlmm bugfix for bug report #2 Apr 2, 2015
.travis.yml added scikit-learn Mar 5, 2015
LICENSE.txt cleaned up data loading/normalization Nov 17, 2014 edited README Mar 5, 2015 new release Apr 2, 2015


Travis PyPI PyPI

A python package implementing warped linear mixed models.

Liner mixed models are widely used in genetics. One of the fundamental assumptions of these models-that the data have a particular distribution (i.e., the noise is Gaussian-distributed)-rarely holds in practice. As a result, standard approaches yield sub-optimal performance, resulting in significant losses in power for GWAS, increased bias in heritability estimation, and reduced accuracy for phenotype predictions.

This repository contains a python implementation of the warped linear mixed model, which automatically learns an optimal "warping function" (or transformation) for the phenotype as it models the data.


Warped Linear Mixed Models for the Genetic Analysis of Transformed Phenotypes. N. Fusi, C. Lippert, N. D. Lawrence and O. Stegle. Nature Communications, 2014


WarpedLMM is available from the python package index.

pip install warpedlmm

Alternatively, if you want access to the code, you can clone this repository.

Getting started

To run WarpedLMM, open a terminal and execute:

python -m warpedlmm SNP_file phenotype_file

SNP_file is a plink BED fileset (e.g. you must have SNP_file.bed, SNP_file.fam, SNP_file.bim). pheno_file is a tab-delimited consisting of 3 columns: individual ID, family ID, phenotype value. For example (look at the test files for more details):

per0	per0	-1.19539
per1	per1	-0.186557
per2	per2	0.561226
per3	per3	-0.808649
per4	per4	0.0214292
per5	per5	-0.471555
per6	per6	-0.456994
per7	per7	-0.740775

It's also possible to specify covariates, write the transformed phenotype values to disk, etc. A list of all the command line options can be accessed using

python -m warpedlmm --help

Citing WarpedLMM

Please cite WarpedLMM in your publications if it helps your research:

         title={Warped linear mixed models for the genetic analysis of transformed phenotypes.},
         author={Fusi, Nicolo and Lippert, Christoph and Lawrence, Neil D and Stegle, Oliver},
         journal={Nature Communications (in press)},

Contacting us

You can submit bug reports using the github issue tracker. If you have any other question please contact: fusi [at] microsoft com, stegle [at] ebi ac uk