eclarke / shortbred-snakemake

Rules to run ShortBRED using Snakemake

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Installation

  1. Clone this directory and ShortBRED into a local folder:

    git clone https://github.com/eclarke/shortbred-snakemake my_project
    cd my_project
    hg clone https://bitbucket.org/biobakery/shortbred
  2. Download the marker genes for CARD from the ShortBRED website:

    wget https://bitbucket.org/biobakery/shortbred/downloads/ShortBRED_CARD_2017_markers.faa.gz
    gunzip ShortBRED_CARD_2017_markers.faa.gz
  3. Get a copy of USEARCH for Linux and install it somewhere. Update the path in the Snakefile to point to wherever you installed usearch. By default it looks in $HOME/bin/usearch.

  4. Create a new Conda environment with Sunbeam (assumes you have Miniconda or similar installed):

    conda create -n my_project -c bioconda snakemake
  5. Copy or symlink your QC'd, decontaminated, paired reads to a folder in this folder (e.g. data_files)

    ln -s /some/other/project/sunbeam_output/assembly/paired data_files
    

Running ShortBRED

  1. Activate your new Conda environment: source activate my_project

  2. Test to make sure Snakemake works correctly (in the folder we made above): snakemake -n. It should output a list of tasks it will perform (but not actually do it yet).

  3. To actually run this, use the command snakemake --use-conda so that it creates a new Python 2.7 environment for ShortBRED. Always use the --use-conda flag when running these rules.

  4. Have fun! I ran this command on PMACS, for reference: snakemake --use-conda --cluster "bsub -n 12" -j100

About

Rules to run ShortBRED using Snakemake


Languages

Language:Python 100.0%