ldsc snakemake workflow

A snakemake workflow to run the ldsc partitioning heritability analysis on predefined genomic regions. https://github.com/bulik/ldsc

LD regression described in this publication:

Bulik-Sullivan, et al. LD Score Regression Distinguishes Confounding from Polygenicity in Genome-Wide Association Studies. Nature Genetics, 2015.

Baseline regions and partition heritability described in this publication:

Finucane, HK, et al. Partitioning heritability by functional annotation using genome-wide association summary statistics. Nature Genetics, 2015

Input

bedfiles with your genomic regions of interest.

Output

tsv file with results from ldsc partitioning heritability for each set of bed files.

All baseline regions appear in the output. The rows with L2_0 in the category column correspond to the enrichment results for the regions in your bedfile.

The results_filename column tells you which bed file that row corresponds to.

Notes

Currently the partitioning h2 is run using the baseline regions from the first ld regression publication as covariates.

Currently only set up to calculate ld using phase 1 of 1000genomes

Usage

To use conda with snakemake, I have had issues using the anaconda modules on the cluster, so I suggest installing your own instance of miniconda

install miniconda

for linux

wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh

I suggest preventing conda from always activating the base environment on start

conda config --set auto_activate_base false

To run the workflow

cd directory_where_you_want_to_save

git clone https://github.com/nrachman/ldsc_snakemake.git

Edit config.yaml to set "workingdir:" and the path to the bed files.

configure any additional options in sm_call.sh (not required)

run

bash sm_call.sh