sGLMM (Sparse Graph-structrued Lasso Mixed model)

Implementation of sGLMM in this paper:

''Ye, Wenting et al. “A Sparse Graph-Structured Lasso Mixed Model for Genetic Association with Confounding Correction.”''

Introduction

sGLMM is a state-of-art model that could extract the genetic variation efficiently from a confounded dataset.

File Structure:

• data/ - we include one simulated sample csv data (500 samples, 2500 SNPs and 25 traits) and a PLINK data
• models/ - main method for sGLMM
• utility/ - other helper files
• runsGLMM.py - main entry point of using sGLMM to work with your own data

How to Use

An Example Command

python runsGLMM.py --snum=50 -n data/mice.plink

this command will run the program and specify the the number of relevant SNPs that the model selects to 50.

python2 runsGLMM.py -n data/simulation --lambda=1 -t csv -m

This command specify the lambda to 1 and run the model once. its input data is csv file in data/ named "simulation". It use simple imputation

Options:

-h, --help          show this help message and exit
-t FILETYPE         choices of input file type (only csv or plink)
-n FILENAME         name of the input file
--lambda=LMBD       The weight of the penalizer. If neither lambda or snum
                    is given, cross validation will be run.
--snum=SNUM         the number of targeted variables the model selects. If
                    neither lambda or snum is given, cross validation will
                    be run.
--threshold=THRESHOLD
                    The threshold to mask the weak genotype relatedness
-q                  Run in quiet mode
-m                  Run without missing genotype imputation

Data Support

• sGLMM currently supports CSV and binary PLINK files.
• Extensions to other data format can be easily implemented through FileReader in utility/dataLoadear. Feel free to contact us for the support of other data format.
• in the

Installation

First, you need to download this repository by using :

git clone https://github.com/YeWenting/sGLMM.git

Then, before you run the program, you need to install following package in your Python, that is :

• Numpy
• Scipy
• Pysnptool

You can install the dependency by using:

pip install -r requirement.txt

Contact

Wenting Ye