detect MNP variants using by Haplotypecaller

Hi!

I used Haplotypecaller 3.1 to detect MNP variants.
java -Xmx6g -Djava.io.tmpdir=$PWD -jar $GATK -R $hg19 -T HaplotypeCaller -I $bamlist --dbsnp $dpsnp135 -o $call/$sample.all.vcf -stand_call_conf 50.0 -stand_emit_conf 10.0 -dcov 200 -nct $nct -A SpanningDeletions -A TandemRepeatAnnotator -A HomopolymerRun -A AlleleBalance -l INFO -baqGOP 30 --max_alternate_alleles 2 -rf BadCigar --minPruning 5

The result had no MNP type varaints, but I can find some continuous snp variants. these snp variants didn't be combined to MNP variants.
Is this normal ? or I really didn't get MNP variants ?

Tagged:

Best Answer

Answers

  • SheilaSheila Broad InstituteMember, Broadie, Moderator

    @zhuwenjuan‌

    Hi. Please post an IGV screenshot of the region and the relevant lines from the VCF file so we can determine what is going on.

    Thanks!

    -Sheila

  • zhuwenjuanzhuwenjuan chinaMember
    edited April 2014

    @Sheila‌ the problem is that none of MNp variants were detected by GATK 3.1. Thanks

    Post edited by zhuwenjuan on
  • zhuwenjuanzhuwenjuan chinaMember
    edited April 2014

    Part of my result:

    1 12987327 rs142575927 G A 762.77 dbSNP AC=2;AF=1.00;AN=2;DB;DP=27;FS=0.000;MLEAC=2;MLEAF=1.00;MQ=25.77;MQ0=0;QD=28.25 GT:AD
    :GQ:PL 1/1:0,27:81:791,81,0

    1 12987349 . G A 1605.77 PASS AC=2;AF=1.00;AN=2;DP=36;FS=0.000;MLEAC=2;MLEAF=1.00;MQ=28.17;MQ0=0;QD=27.93 GT:AD:GQ:PL
    1/1:0,36:99:1634,111,0

    1 12987350 . A T 1605.77 PASS AC=2;AF=1.00;AN=2;DP=37;FS=0.000;MLEAC=2;MLEAF=1.00;MQ=28.04;MQ0=0;QD=33.89 GT:AD:GQ:PL
    1/1:0,37:99:1634,111,0

    1 12987602 rs76625528 T C 797.77 dbSNP AC=2;AF=1.00;AN=2;DB;DP=28;FS=0.000;MLEAC=2;MLEAF=1.00;MQ=28.24;MQ0=0;QD=28.49 GT:AD:GQ:PL 1/1:0,28:84:826,84,0

    It looks like I can't upload the IGV picture. this MNP has high quality.

    Post edited by zhuwenjuan on
  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie

    Just to be clear -- what you're expecting is for positions 1:12987349 and 1:12987350 to be emitted together as a MNP with alleles GA/AT, correct?

  • zhuwenjuanzhuwenjuan chinaMember

    yes, 1:12987349 and 1:12987350 , now just continuous snp variants ,not combine together. If I care about MNP, I cann't directly extract this variants.

  • Is the --mergeVariantsViaLD still a thing? I recently ran into the issue of a 3bp substitution being called as three SNPs. I did not think MNPs were an issue for haplotype caller. I do not see the --mergeVariantsViaLD parameter in either HaplotypeCaller or GenotypeGVCFs documentation.

  • SheilaSheila Broad InstituteMember, Broadie, Moderator

    @bwubb
    Hi,

    No, that is not an option in HaplotypeCaller. Have a look at this thread. You can use ReadBackedPhasing to merge MNPs.

    -Sheila

Sign In or Register to comment.