Open sorelfitzgibbon opened 1 year ago
This is interesting! I was wondering about the accuracy compared to call-mtSNV!
Currently the pipeline is using normal Mutect2 on chrM. Should chrM either be removed from the calling, or should call-mtSNV be pulled in to make the calls?
Just a quick thought, if we were to add Mutect2 mitochondrial calling, we'd want to add it to call-mtSNV since we have a pipeline set up for that specific purpose. I'm not sure whether we want to remove it here or keep it as is and advise users to use call-mtSNV calls for mitochondrial variants.
Just a quick thought, if we were to add Mutect2 mitochondrial calling, we'd want to add it to call-mtSNV since we have a pipeline set up for that specific purpose. I'm not sure whether we want to remove it here or keep it as is and advise users to use call-mtSNV calls for mitochondrial variants.
Yup, that makes sense. Also, this task is considered low priority for now.
Mutect2 has a mitochondrial mode that deals with issues specific to mitochondrial variant calling, discussed here: https://github.com/broadinstitute/gatk-docs/blob/master/blog-2012-to-2019/2019-03-05-New!_Mitochondrial_Analysis_with_Mutect2.md