silly-gat-kay

Repository for Docker image astrabert/silly-gat-kay info and changelog

Information on the image

Variant calling and statistical analysis of population genomics data requires a certain number of tools, which can differ for functioning and version between machines and platforms. This is why we created a virtual environment managed through Docker, based on broadinstitute/gatk: you can run the image inside a container, which will produce an isolated machine on which you can carry out your analysis, without conflicts.

The image is easily accessible through;

sudo docker pull astrabert/silly-gat-kay:latest

You should run it in interactive mode, possibily mounting your local file system into the container if you want to access data from your real machine inside the virtual one.

sudo docker run -v /path/to/your/project:/gatk_modified/drosophila-project -it astrabert/silly-gat-kay

Otherwise, adjust the path to the data you want to mount inside your Docker image in env under USERDATA_PATH and then simply run:

docker compose up -d
docker exec -it $(sudo docker ps -qf "name=drosophila_project_env") /bin/bash

Or, if you are on Linux/macOS:

bash compose.sh

You'll be put inside the container (a semi-isolated virtual machine) in which our Docker image is up and running, and you'll find all your mounted data under /gatk_modified/userdata.

The virtual machine is based on broadinstitute/gatk and provides a fully equipped environment for the analysis, divided into three conda envs:

• gatk_modified: contains all GATK dependencies, but also bwa-mem2 and cutadapt
• picard: contains picard-slim a lighter implementation of Picard which does not require r-base to run
• platypus: contains platypus-variants, a SNP-calling focused conda wrapper for Platypus
• R: contains r-base platform for data analysis in R and the library poolfstat
• python_deps: contains python3.11 and various libraries like pandas, biopython, pyyaml.
• freebayes-env: contains freebayes to run variant calling.

To activate the conda envs and deactivate them, run:

## activate
source activate platypus
## deactivate
conda deactivate

Apart from these conda envs, it also contains other tools like cmake, the basic text editor micro, the multi-threading gzipper pigz and the SRA toolkit. To make SRA-toolkit CLI work, you should run:

cd /gatk_modified
export PATH=$PATH:$PWD/sratoolkit.3.1.1-ubuntu64/bin
## check if it was successful
fastq-dump -h