gatk-sv
gatk-sv copied to clipboard
Hail VCF concatenation in Terra
In the MakeCohortVcf
module, we added an option to use Hail to concatenate VCFs on a Spark cluster, which was greatly faster (wall time wise) than single-threaded methods. However, Spark clusters cannot currently be created within workflows on Terra.
@cwhelan suggested trying to use Hail locally on a large VM as a workaround. We should test this against bcftools concat
in terms of wall time, and, if beneficial, add as another option to the module.