Reproduce the final steps

You can download intermediate files at Zenodo (10.5281/zenodo.1488638):

Put both of these files without gunzipping in the out/ directory. Download the CIS-BP TF information:

Unzip this file. Edit the script scripts/create_clustered_pfm.py and change the following line to the right location:

tf_info = "/data/cis-bp-1.02/TF_Information_all_motifs.txt"

By running the script you will re-create the clustered pfm file.

python scripts/create_clustered_pfm.py

Run the full pipeline

These steps will reproduce the complete pipeline.

The file config.yaml contains the configuration:

pwm:  "data/20181017.all.pfm"
chunksize: 100

To run this pipeline you will have to install bioconda. Install snakemake using conda:

conda install snakemake

Dry run:

snakemake -n

Full run using 12 cores:

snakemake --use-conda --cores 12

See cluster.json and the script submit_snakemake.sh for an example on how to run the workflow on a cluster (SLURM in this case).