All dependencies for the primary workflow are installed via conda. To set up your environment first make sure you channel priorities are set correctly according to the bioconda documentation:
conda config --add channels defaults
conda config --add channels bioconda
conda config --add channels conda-forge
conda config --set channel_priority strict
Then install the conda environment used by the workflow with these commands:
git clone https://github.com/fiberseq/fiberseq-smk
cd fiberseq-smk
conda create -n fiberseq-smk
mamba env update -n fiberseq-smk --file workflow/envs/env.yml
conda activate fiberseq-smk
If you want to use the ipdSummary
pipeline you will need to install some additional dependencies available through SMRTLINK
.
# Add the SMRTlink tools to your path
export PATH=$PATH:/path/to/smrtlink/tools
If you are on hyak
you can add our copy to your path by adding this to your .bashrc
:
PATH=$PATH:/gscratch/stergachislab/install_dir/smrtlink/smrtcmds/bin/