This is a Snakemake pipeline to convert genetic variant information in VCF format to a format used for inference in IMa3. Essentially, this is an automated application of the Popgen Pipeline Platform for a specific application.
There are no specific installation steps required for this pipeline other than having the required dependencies. One can begin by cloning the repository to their local machine:
git clone [email protected]:Chris1221/IMa3_Input.git
cd IMa3_Input
To install all of the required dependencies, you may use the conda environment provided.
Assuming that you have conda installed locally, activate the environment by using the provided file.
conda env create --file=env.yml
conda activate ima3_input
For basic usage, the only file you need to edit is config.yaml
. This file is read by the pipeline and steps are performed according to your personal set up.
The required options that you need to edit are:
input
: The file path to your VCF.analysis_name
: A unique identifier for this run of the pipeline.
Edit the configuration file to give your options:
- name:
+ name: /path/to/your/file.vcf.gz
- analysis_name:
+ analysis_name: my_first_run
Run the pipeline by using either the input
or run
directives.
snakemake input -j1 # -j1 means to use a single core
snakemake run -j20
The inference_parameters
section of the configuration file is flexible. If you include an option here, it will be given as a flag to IMa3
.
IMa3
is not easily installable, so you will have to go through the process yourself by following the official guide. Tell the pipeline where your installation lives by editing the-ima3_path
variable in theinference_paramters
section of the config file (the leading dash is to allow for two dashes when executing the script).- You can run a test version of the pipeline on 10k variants from SGDP by not editing the configuration file. This is a good way to check that all dependencies have installed correctly.
- A description of IMa3's extensive list of options is beyond the scope of this project, please check out the documentation PDF (and good luck...).