gatk/haplotypecaller-gvcf-gatk4 hangs up on Firecloud

lelagina

Hello,

The gatk/haplotypecaller-gvcf-gatk4 workflow on Firecloud starts and hangs without producing a workflow log, JES log, stderr or stdout file.

I noticed that in runtime section for each task in wdl script for this workflow it doesn't use ${} for pass in parameters.

_runtime {
docker: docker_image
memory: mem_size
disks: "local-disk " + disk_size + " HDD"
}
_
instead of

runtime {
docker: ${docker_image}
memory: ${mem_size}
disks: "local-disk " + disk_size + " HDD"
}

Is this correct?

Thank you.

bshifaw

Yes, there is a featured workspace using the method on b37(hg19) reference data and test bam.
A workflow may slow when pulling docker images from docker hub instead of the images in Google Cloud Repository, where are you pulling your docker image from?

bshifaw

The parameters should still work using that format. Were you able to successfully run the method using the test dataset and the method configuration in the featured workspace?

lelagina

Yes, I was able to run it. I see that in that workspace it is run on hg38 aligned bams. Is it possible to run Haplotype Caller method (haplotypecaller-gvcf-gatk4) on hg19 aligned bams? Method description doesn't specify it.

bshifaw

Yes, there is a featured workspace using the method on b37(hg19) reference data and test bam.
A workflow may slow when pulling docker images from docker hub instead of the images in Google Cloud Repository, where are you pulling your docker image from?