search

broadinstitute / gatk

Official code repository for GATK versions 4 and up
https://software.broadinstitute.org/gatk
Other
1.68k stars 587 forks source link

Potential CNV? #3149

Open chandrans opened 7 years ago

chandrans commented 7 years ago

Bug Report

Affected tool(s)

HaplotypeCaller

Affected version(s)

latest GATK4

Description

A user reported a verified 50 bp insertion is not being called with HaplotypeCaller. On closer inspection, I noticed the coverage is really odd in the region. Original BAM:

screen shot 2017-06-16 at 4 21 23 pm

bamout:

screen shot 2017-06-16 at 4 22 10 pm

Steps to reproduce

The files are here: /humgen/gsa-scr1/schandra/asaki_InsertionNotCalled

Command: `./gatk-launch HaplotypeCaller -R human_g1k_v37.fasta -I GW

screen shot 2017-06-16 at 4 22 10 pm

I127.rs1799752.bam -O Sheila.HaplotypeCaller.vcf -bamout Sheila.HaplotypeCaller.bam`

Expected behavior

Is this insertion supposed to be called? Need clarification from Methods.

This Issue was generated from your forums

droazen commented 7 years ago

@chandrans HaplotypeCaller (even in GATK3) can't usually call large indels. Does GATK3 HaplotypeCaller make this call?

I'll also add: GATK4's HaplotypeCaller is beta, and not yet ready for production use. It does not currently match the output of the GATK3 HaplotypeCaller.

chandrans commented 7 years ago

Hi David. This was not called in GATK3 either. It is more a question of whether this should be called by HC because the user says it is a verified indel. We should be calling verified true indels. The user says it is 50bp, and HC has called things larger than that before.

I am not sure if this is a bug, but more asking if this is a potential CNV from the looks of the coverage? Or, what kind of args the user can provide to get the "indel" called?