Installation

DRAKKAR can be installed directly from the GitHub repository or accessed as an HPC module on supported systems (e.g., Mjolnir). The core workflow is executed with Snakemake and uses Conda environments for tool dependencies.

Install From GitHub

$ pip install git+https://github.com/alberdilab/drakkar.git

If you are working in a managed HPC environment, prefer using a dedicated Conda environment:

$ conda create -n drakkar python=3.12
$ conda activate drakkar
$ pip install git+https://github.com/alberdilab/drakkar.git

Use on Mjolnir (module)

If your cluster provides a DRAKKAR module, load it before running any command:

$ module load drakkar/<version>

Verify installation

$ drakkar --help

Optional: pre-create environments

You can pre-create Conda environments for all workflows (useful on clusters where environment creation during a run is slow):

$ drakkar environments --profile slurm

If you want to use a shared environment directory, pass -e/--env_path to commands or set it in your workflow config.