Notice:
If you happen to see a question you know the answer to, please do chime in and help your fellow community members. We encourage our fourm members to be more involved, jump in and help out your fellow researchers with their questions. GATK forum is a community forum and helping each other with using GATK tools and research is the cornerstone of our success as a genomics research community.We appreciate your help!

Test-drive the GATK tools and Best Practices pipelines on Terra


Check out this blog post to learn how you can get started with GATK and try out the pipelines in preconfigured workspaces (with a user-friendly interface!) without having to install anything.

Is normal-artifact-lod too stringent in FilterMutectCalls

darioberdariober Cambridge UKMember

Hi- Using FilterMutectCalls from gatk 4.0.4.0, I see that the --normal-artifact-lod option is set to 0.0 and I wonder whether this is too stringent.

The variant below has been flagged as artifact_in_normal for having N_ART_LOD=0.953 which results from 1 mismatching read out of 45 total reads (I checked the alignments and there are no other mismatching reads). Considering that the tumour has loads of supporting reads (REF, ALT: 44, 42) this variant would be rejected despite being a likely true positive.

Maybe one should also consider the difference between N_ART_LOD and/or TLOD and/or NLOD?

chr12 25245351 . C A . artifact_in_normal DP=135;ECNT=1;NLOD=9.27;N_ART_LOD=0.953;POP_AF=1e-05;P_CONTAM=0;P_GERMLINE=-7.979;TLOD=149.47;CSQ=A|...|HGNC:6407| GT:AD:AF:F1R2:F2R1:MBQ:MFRL:MMQ:MPOS:SA_MAP_AF:SA_POST_PROB 0/1:44,42:0.49:11,13:33,29:38:167,154:60:15:0.444,0.475,0.488:0.038,0.012,0.949 0/0:44,1:0.043:12,0:32,1:38:169,135:60:20:.:.

Any thoughts much appreciated!

Dario

Best Answer

  • darioberdariober Cambridge UK
    Accepted Answer

    Hi Sheila - I just tested with v4.0.8.1 and the variant is marked as PASS. So, yes, the latest version fixes it! Thanks a lot and apologies for not trying this before!

Answers

  • SheilaSheila Broad InstituteMember, Broadie admin

    @dariober
    Hi Dario,

    This is a bit odd. Can you post some IGV screenshots of the BAM file and bamout file? I just want to see if there are more reads supporting the alt allele in the normal than are represented in the VCF.

    Thanks,
    Sheila

  • darioberdariober Cambridge UKMember

    Hi Sheila, thank you for your reply and apologies I couldn't reply earlier.

    In the tar archive attached here I put a screenshot of the reads (sorry, it's not IGV though) and the bam files with the reads in the region chr12:25245226-25245475 (X32N_sc_X32N.bam is the normal). As reported in the vcf, it seems to me there is just one read in the normal having an A at chr12:25245351.

    Thank you!

    Dario

  • hello @Sheila !

    right now we are facing the same problem with 4 WES of same sample - all are different sites of same tumor:

    chr20 2798332 . A AGTGG . PASS DP=218;ECNT=1;NLOD=31.42;N_ART_LOD=0.309;POP_AF=2.500e-06;P_CONTAM=1.170e-20;P_GERMLINE=-3.739e+01;RPA=1,2;RU=GTGG;STR;TLOD=32.03 GT:AD:AF:F1R2:F2R1:MBQ:MFRL:MMQ:MPOS:OBAM:OBAMRC:SA_MAP_AF:SA_POST_PRO0/1:53,10:0.220:22,7:31,3:34,35:193,188:60:34:false:false:0.131,0.141,0.159:0.015,0.013,0.972 0/0:118,1:0.084:51,0:67,1:33,31:182,130:60:4:false:false

    chr20 2798332 . A AGTGG . artifact_in_normal DP=240;ECNT=1;NLOD=27.86;N_ART_LOD=3.00;POP_AF=2.500e-06;P_CONTAM=0.00;P_GERMLINE=-3.881e+01;RPA=1,2;RU=GTGG;STR;TLOD=29.92 GT:AD:AF:F1R2:F2R1:MBQ:MFRL:MMQ:MPOS:OBAM:OBAMRC:SA_MAP_AF:SA_POST_PROB 0/1:74,10:0.166:37,6:37,4:33,33:192,229:60:26:false:false:0.101,0.101,0.119:0.011,0.012,0.977 0/0:120,2:0.089:52,0:68,2:33,32:182,108:60:4:false:false

    chr20 2798332 . A AGTGG . artifact_in_normal DP=245;ECNT=1;NLOD=27.85;N_ART_LOD=3.00;POP_AF=2.500e-06;P_CONTAM=1.708e-20;P_GERMLINE=-4.173e+01;RPA=1,2;RU=GTGG;STR;TLOD=28.47 GT:AD:AF:F1R2:F2R1:MBQ:MFRL:MMQ:MPOS:OBAM:OBAMRC:SA_MAP_AF:SA_POST_PROB 0/1:85,10:0.132:33,5:52,5:33,33:190,225:60:33:false:false:0.071,0.101,0.105:0.025,5.848e-03,0.969 0/0:120,2:0.086:52,0:68,2:33,32:182,108:60:4:false:false

    chr20 2798332 . A AGTGG . artifact_in_normal DP=241;ECNT=1;NLOD=27.87;N_ART_LOD=3.00;POP_AF=2.500e-06;P_CONTAM=8.214e-14;P_GERMLINE=-3.793e+01;RPA=1,2;RU=GTGG;STR;TLOD=28.83 GT:AD:AF:F1R2:F2R1:MBQ:MFRL:MMQ:MPOS:OBAM:OBAMRC:SA_MAP_AF:SA_POST_PROB 0/1:71,10:0.184:35,4:36,6:32,32:179,159:60:40:false:false:0.111,0.101,0.123:8.733e-03,0.016,0.975 0/0:120,2:0.089:52,0:68,2:33,32:182,108:60:4:false:false

    our m2 followed https://software.broadinstitute.org/gatk/documentation/article?id=11136 except we had no PoN

    Bests,

    Sergey

  • SheilaSheila Broad InstituteMember, Broadie admin

    @dariober @sergey_ko13
    Hi Dario and Sergey,

    I am not sure why this is happening. I don't think it should happen. Before I check with the developer, can you confirm that this happens in the latest version?

    Thanks,
    Sheila

  • darioberdariober Cambridge UKMember
    Accepted Answer

    Hi Sheila - I just tested with v4.0.8.1 and the variant is marked as PASS. So, yes, the latest version fixes it! Thanks a lot and apologies for not trying this before!

  • Hi @Sheila
    we just tried 4.0.8.1. It happened again. So 4 samples of one Tumor with same Normal. 1 has PASS and 3 marked with artifact_in_normal

    Thank You,
    Sergey

  • UPD:
    we have run m2 4.0.8.1 from the very beginning (not only FilterMutectCalls) - now all 4 samples have artifact_in_normal in that position. It's easier to accept that all are filtered out by criteria - because all share same normal sample

    no need to bother developers with that yet.
    Thank You!

    Sergey

Sign In or Register to comment.