Variant calling + processing pipline using GATK based in Snakemake
- Use branch maser for calling with GATK 4
- Use branch gatk37 for calling with GATK 3.7
The GATK4 calling currently does not work due to an issue with how gatk SplitNCigarReads parses its arguments
source activate [gatk,gatk37] # activate conda enviornment
See here to create the conda environment
snakemake --dag --configfile "XXX-config.yaml" | dot -Tpng > "XXX-worflow.png"
snakemake --configfile "XXX-config.yaml" --cores "N"
Additional parameters can be passed to the Haplotype caller step by way of the hcArgs variable in the config file. For example to restrict genotyping to specific positions on the genome we would add the following line to the config file. (ip provides padding arond each site, recommended by GATK site.)
hcArgs: "-L my_variants.vcf -ip 100"
- Fix GATK 4 calls so it works - is this a bug or implementation issues?
This current process is very much not ideal, but the conda package for gatk4 is not standalone so is unavoidable.
- Clone the gatk 4 repo
Check if this version is that latest.
wget https://github.com/broadinstitute/gatk/releases/download/4.0.2.1/gatk-4.0.2.1.zip
unzip gatk-4.0.2.1.zip
cd gatk-4.0.2.1.zip
- Make the conda environment
conda env create -n gatk4 -f gatkcondaenv.yml
- Add Snakemake and java 1.8 to the conda environment (graphviz is for making snakemake DAGs)
source ~/miniconda/bin/activate gatk4
conda install -c bioconda java-jdk snakemake graphviz
Then update the gatkPath variable in config.yaml with the path to the gatk executable from the gatk downlaod.
To use the GATK 3.7 branch follow these instructions to set up the conda environment.
conda create -n gatk37 gatk=3.7
source activate gatk37
Then download version 3.7 of GATK from https://software.broadinstitute.org/gatk/download/archive
gatk-register "/path/to/gatk-XX-tar.bz2"
conda install -c bioconda java-jdk snakemake graphviz