LEM

A pipeline to extract leading edge metagene from GSEA results

Install dependency

pip install -r requirements.txt

Run scripts

The project folder structure should be the same as this repo, put code in the 'Code', data in 'Data', and add an empty folder called Result.

1. Estimate the number of ranks

To get the help information just run

python 1-nmfRankEstimation.py --help 

To run the script

python 1-nmfRankEstimation.py -f <File Name> -n <Number of Runs> -s <Start Rank> -e <End Rank>

2. Calculate the W matrix

To get the help information just run

python 2-nmfComputeW.py --help 

To run the script

python 2-nmfComputeW.py -r <Number of Runs>

3. Calcualte cutoff

Rscript 3-calculate.cutoff.R <File Name>

The file used here is produced in the step 2

4. Assign membership

Rscript 4-NMF.Membership.Assign.R <File Name> <Threshold>

The file used here is the same as step 3 and threshold is decided based on the calculation in step 3