Skip to content

Installation

1. Install conda (skip this step if you already have conda installed)

If you do not already have conda installed, you can check out the documentation here. We recommend you install miniforge

2. Create and install bohra

mamba (or conda) create -n bohra -c bioconda bohra

3. Install dependencies

conda activate bohra
bohra deps install

4. Set environment variables (optional)

kraken2 You will need to ensure that you have a kraken2 database available. If you do not set this as an environment variable you will need to provide the path to the kraken2 database you wish to use at runtime. IMPORTANT - if you do not have a kraken2 database set species specific functions of bohra will not work.

You may already have a kraken2 database environment variable set on your system. To check this 

echo $KRAKEN2_DEFAULT_DB
If this outputs a path - this is the default kraken2 database.

If this is empty or if you wish to use a different database you can set it at runtime OR set the path in your bohra environment.

conda env config vars set KRAKEN2_DEFAULT_DB=/path/to/kraken2_db
it will tell you to deactivate and re-activate your environment

conda deactivate bohra
conda activate bohra

Now you can test that the environment variable has been set 

echo $KRAKEN2_DEFAULT_DB
should print 
/path/to/kraken2_db
mlst

If you have your own mlst database you can provide this to bohra. Similar to setting the KRAKEN2_DEFAULT_DB.

You may already have a kraken2 database environment variable set on your system. To check this 

echo $MLST_DBDIR
If this outputs a path - this is the default mlst database.

If this is empty or if you wish to use a different database you can set it at runtime OR set the path in your bohra environment.

conda env config vars set MLST_DBDIR=/path/to/mlst_db
it will tell you to deactivate and re-activate your environment

conda deactivate bohra
conda activate bohra

Now you can test that the environment variable has been set 

echo $MLST_DBDIR
should print 
/path/to/mlst_db

5. Testing your installation

If you would like to make sure that all the dependencies are installed properly and that the pipeline will run on your system you can run

bohra test --cpus X

(where X is the number of cpus you want to use)

This will download some fastq files and run the full pipeline.

Install from source

Setup the bohra environment and install the pipeline co-ordinator

  1. Clone the bohra repository
git clone https://github.com/MDU-PHL/bohra.git

. Create the environment and install bohra

cd bohra
conda env create -f environment.yml -n bohra
conda activate bohra
pip3 install .

Install dependencies

This step will setup conda environments under the path/to/conda/envs/bohra(depending on how you have configured your conda installation).

  1. Activate your environment from step 2 above 

    conda activate bohra

  2. Run the bohra dependency installation

bohra  deps install

Updating bohra

Updates to the bohra pipeline will often include 2 steps/

  1. Update bohra env
conda update -n bohra --all

This will update the bohra pipeline and running scripts. But not any dependencies

  1. Update any changed dependencies
conda activate bohra
bohra deps update

This will update any dependencies that have pinned versions or any new dependencies and should be run each time you update bohra.

Other update options

You may also update a specific tool or dependency environment. This should be done with caution as updates that have not been adeqautely incorporated into the pipeline may cause unexpected behaviour or breakages. 

conda activate bohra
bohra deps update --tool <env_name> --force

If you would like another tool or different versions please leave an issue and we will endeavour to facillitate it.

If you need to re-install your dependencies for some reason you can run

conda activate bohra
bohra deps update --force
This will overwrite all existing dependencies and may take some time.