We've moved!
This site is now read-only. You can find our new documentation site and support forum for posting questions here.
Be sure to read our welcome blog!

No MQ annotation from Haplotype Caller in version 3.5 but does work in 3.4

I upgraded to GATK 3.5 to use MuTect2 and it works great! However, I'm now using the updated .jar in my germline variant calling pipeline and there's some issues. I only noticed it when I went to run VQSR, and the INDELs went through but not the SNPs. I got an error in the SNP log about the MQ annotation. It seems to be completely absent in the haplotype caller output (the .g.vcf's). I tried forcing it with -A RMSMappingQuality but it did not affect the output. However it works fine if I go back to 3.4

Command Line for 3.4:
java -Xmx4g -jar /REDACTED/tools/gatk_3.4.0/GenomeAnalysisTK.jar -T HaplotypeCaller -R /REDACTED/resources/b37/gatkBundle/human_g1k_v37_decoy.fasta -I /REDACTED/melanoma/bam/recal/PFM-32.ra.rc.bam -L /REDACTED/resources/b37/targetRegions/nexteraRapidCapture_expandedExome/1.bed -ERC GVCF -o /REDACTED/melanoma/variantCall/gatk3.4/gvcf/PFM-32/PFM-32_chr1.g.vcf -A RMSMappingQuality

Command Line for 3.5:
java -Xmx4g -jar /REDACTED/tools/gatk_3.5.0/GenomeAnalysisTK.jar -T HaplotypeCaller -R /REDACTED/resources/b37/gatkBundle/human_g1k_v37_decoy.fasta -I /REDACTED/melanoma/bam/recal/PFM-32.ra.rc.bam -L /REDACTED/resources/b37/targetRegions/nexteraRapidCapture_expandedExome/1.bed -ERC GVCF -o /REDACTED/melanoma/variantCall/gvcf/PFM-32/PFM-32_chr1.g.vcf -A RMSMappingQuality

And here's a single line from the output of each at the same position

1 1431520 . C A,<NON_REF> 118.77 . BaseQRankSum=0.452;DP=14;MLEAC=1,0;MLEAF=0.500,0.00;MQ=60.00;MQRankSum=-1.485;ReadPosRankSum=-0.452 GT:AD:GQ:PL:SB 0/1:8,6,0:99:147,0,187,170,205,375:4,4,2,4

1 1431520 . C A,<NON_REF> 118.77 . BaseQRankSum=0.710;ClippingRankSum=0.452;DP=14;ExcessHet=3.0103;MLEAC=1,0;MLEAF=0.500,0.00;MQRankSum=0.452;RAW_MQ=50400.00;ReadPosRankSum=-1.614 GT:AD:DP:GQ:PL:SB 0/1:8,6,0:14:99:147,0,187,170,205,375:4,4,2,4

I didn't notice anything relevant in the 3.5 release notes except for something about VQSR and MQ jittering which leads me to believe that MQ is indeed a valid annotation still, but how can I get it?

Thanks in advance

Issue · Github
by Sheila

Issue Number
Last Updated
Closed By

Best Answer


  • c.martinez11c.martinez11 NYCMember

    I was following the best practices but your comment led to the correct answer. I looked more closely at my GenotypeGVCFs script and found that it was still using the 3.4 jar and not 3.5, so MQ was never calculated. Thank you so much!

Sign In or Register to comment.