VariantRecalibrator erro: Bad input: Found annotations with zero variance.

Hi,
I'm trying to run GATK recalibration pipeline to my data:

java -jar /opt/bin/GenomeAnalysisTK-3.2-0/GenomeAnalysisTK.jar -T VariantRecalibrator -R ../reference/chrall.fa -input output.vcf \ -resource:hapmap,known=false,training=true,truth=true,prior=15.0 ../hapmap_3.3.hg18.vcf \ -resource:omni,known=false,training=true,truth=true,prior=12.0 ../1000G_omni2.5.hg19.sites.vcf \ -resource:1000G,known=false,training=true,truth=false,prior=10.0 ../1000G_phase1.indels.hg19.sites.vcf \ -resource:dbsnp,known=true,training=false,truth=false,prior=2.0 ../dbsnp_138.hg19.vcf \ -an DP \ -an QD \ -an FS \ -an MQRankSum \ -an ReadPosRankSum -mode SNP \ -tranche 100.0 -tranche 99.9 -tranche 99.0 -tranche 90.0 \ -recalFile recalibrate_SNP.recal -tranchesFile recalibrate_SNP.tranches -rscriptFile recalibrate_SNP_plots.R

But I'm getting the folling error:

**##### ERROR

ERROR MESSAGE: Bad input: Found annotations with zero variance. They must be excluded before proceeding.
ERROR ------------------------------------------------------------------------------------------**

Can anyone tell me what I'm doing wrong?
Thanks,

Rômulo

Answers

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie

    Hi Romulo,

    Can you tell us a little more about how the data was generated? How did you perform the variant calling?

  • romulucromuluc BrazilMember

    Dear Geraldine,

    We have a run with a dozen of samples... I just run calling to one, using in input all chromossomes as reference, but a put -L chr11 to focus result to get a snp which I'm trying to recover:

    java -jar /opt/bin/GenomeAnalysisTK-3.2-0/GenomeAnalysisTK.jar -T HaplotypeCaller -I B1.bam -R ../chrall.fa -L chr11 --genotyping_mode DISCOVERY -stand_emit_conf 10 -stand_call_conf 30 -o output.vcf

    Something is wrong ?
    Thanks!

    Rômulo

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie

    Oh I see. You can't run Variant Recalibration on only a chromosome's worth of data because that does not provide enough information to the program. The typical requirement for VQSR is either one whole-genome sample or ~30 exomes.

  • romulucromuluc BrazilMember

    Geraldine,

    So I have to run the variant calling without -L parameter?

    Thanks!
    Rômulo

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie

    Yes, if your samples are whole-genome. If they are exomes, you should use -L to provide the exome capture targets list.

Sign In or Register to comment.