Mutect2 with lastest nightly build problem

ac67479ac67479 AustinMember

Hi,

I am trying to run the variant calling step for the first time as a test. As I don't have any matched normal, I have only my tumour .bam file and a PoN composed of only 2 samples (for now).

So I combined my two .vcf normals to create a PoN and a launched MuTect2 like that :

java -jar /analysis/GATK/GenomeAnalysisTK-3.6/GenomeAnalysisTK.jar \
-T MuTect2 \
-R /projects/acoudray/star/ref_genome/hg38_gatk/GRCh38.primary_assembly.genome.fa \
-I:tumor /projects/acoudray/star/RNAseq/BaseRecalibration/rnaseq_added_sorted_dedup_split_recal.bam \
-PON MuTect2_PON_2samples.vcf \
--dbsnp /projects/acoudray/star/ref_genome/dbsnp_b147_hg38.vcf \
-o MuTect2_output_SD01_RNAseq_PON2samples.vcf

and I got some strange warning during the compilation :

WARN 18:01:55,963 SomaticGenotypingEngine - At Locus chrchr1:8685652, we detected that variant context had alleles that not in PRALM. VC alleles = [C, T], PRALM alleles = []
WARN 18:01:56,283 SomaticGenotypingEngine - At Locus chrchr1:8717782, we detected that variant context had alleles that not in PRALM. VC alleles = [A, G
], PRALM alleles = []
WARN 18:01:56,458 SomaticGenotypingEngine - At Locus chrchr1:8733981, we detected that variant context had alleles that not in PRALM. VC alleles = [C, T], PRALM alleles = []
WARN 18:01:57,860 SomaticGenotypingEngine - At Locus chrchr1:8808360, we detected that variant context had alleles that not in PRALM. VC alleles = [A, C
], PRALM alleles = []
INFO 18:02:03,694 ProgressMeter - chr1:8861373 0.0 5.5 m 545.7 w 0.3% 32.1 h 32.0 h
INFO 18:03:03,696 ProgressMeter - chr1:8865651 0.0 6.5 m 644.9 w 0.3% 37.9 h 37.8 h

and it seems that nothing is written in the output.

Any idea from where it comes from ?

Note : I just installed the nightly build for the 2 last steps (combining .vcf files and variant calling with mutect2). The normal-only mode to generate .vcf file in artifact_detection_mode wasn't done with the latest nightly build but with GATKv3.6 without any nightly build.

Thanks a lot !

Alex

Tagged:

Best Answers

Answers

  • ac67479ac67479 AustinMember

    my .vcf files seem to bear proper informations from the variants detected in artifact_detection_mode

  • ac67479ac67479 AustinMember

    I rectify : The VFC file I made without the nightly build seems good. But I tried to re-produce them with the nightly build and I obtained the same warning as above, and nothing is written on the VCF PON file !!

  • SheilaSheila Broad InstituteMember, Broadie, Moderator admin

    @ac67479
    Hi,

    Okay, I am a little confused. Can you try running with the latest nightly build in artifact detection mode? Don't include the PON and dbSNP files for now. Let's see if you get a proper output with just the basic arguments.

    Thanks,
    Sheila

  • ac67479ac67479 AustinMember

    Hello ! Finally even with these warnings the .vcf files was created and the output look like that:
    chr1 14699 . C G . PASS ECNT=1;HCNT=2;MAX_ED=.;MIN_ED=.;TLOD=8.75 GT:AD:AF:ALT_F1R2:ALT_F2R1:FOXOG:QSS:REF_F1R2:REF_F2R1 0/1:1,3:0.750:0:3:1
    .00:20,83:0:1
    Is it what I am expecting ?

  • ac67479ac67479 AustinMember

    I am doing right now the artifact detection with the nightly build on chr1 ! I'll post the results here as soon as I can :) Thanks a lot !

  • ac67479ac67479 AustinMember

    Yes it was produced with the latest nightly build ! I also tried the artifact detection mode with the latest nightly build and obtained this :

    chr1 191870 . C A . t_lod_fstar ECNT=2;HCNT=1;MAX_ED=8;MIN_ED=8;TLOD=4.50 GT:AD:AF:ALT_F1R2:ALT_F2R1:FOXOG:PGT:PID:QSS:REF_F1R2:REF_F2R1 0
    /1:0,2:1.00:0:2:1.00:0|1:191870_C_A:0,60:0:0

    which was similar to the one I've got without the nightly build. However, I still get this kind of warnings :

    WARN 18:01:57,860 SomaticGenotypingEngine - At Locus chrchr1:8808360, we detected that variant context had alleles that not in PRALM. VC alleles = [A, C], PRALM alleles = []

    So if you think my vcf looks good, I'll try to understand the content of the vcf ! Thank you so much

  • ac67479ac67479 AustinMember

    Anyway, I am still wondering what is this PRALM ?

  • murphycjmurphycj New York CityMember

    Hello, I' am getting this same warning message and want to find a way around it. I' am currently working with MiSeq data (so my coverage is in the tens of thousands). As a result, I've noticed that MuTect is not emitting sites which are clearly somatic (as determined by other mutation calling algorithms). How can I get MuTect to emit everything?

  • SheilaSheila Broad InstituteMember, Broadie, Moderator admin

    @murphycj
    Hi,

    Can you please post the exact command and version you are running? Please also post some IGV screenshots of sites you think should be called, but are not.

    Thanks,
    Sheila

  • jfpolonijfpoloni BrazilMember

    Hi,
    I’m getting the same warning and I can’t figure out what is the problem.
    I’m using GATK v3.7-0-gcfedb67 and my command line is:

    java -Xmx4g -jar $GATK -T MuTect2 \
    -R $Reference/Homo_sapiens.GRCh38.dna.primary_assembly.fa \
    -I:tumor $WD/SRR975551tumor.recal_reads.bam \
    -I:normal $WD/SRR975569normal.recal_reads.bam \
    --dbsnp $Reference/dbsnp_146.hg38.vcf.gz \
    --cosmic $Reference/CosmicCodingMuts.vcf.gz \
    --cosmic $Reference/CosmicNonCodingVariants.vcf.gz \
    -o $WD/Mutect2/51_69output_PTxN.vcf

    Thank you!

  • murphycjmurphycj New York CityMember

    @Sheila

    Thanks for the reply, but I haven't been able to reproduce the issue I raised. It was just something I observed through some brief data exploration and must have been mistaken.

  • SheilaSheila Broad InstituteMember, Broadie, Moderator admin

    @jfpoloni
    Hi,

    Are you talking about the PRALM warnings? Those are nothing to worry about, as I mentioned above.

    If not, please post the WARN message you are talking about? Or, is your issue the same as @murphycj ?

    Thanks,
    Sheila

  • jfpolonijfpoloni BrazilMember

    Hello @Sheila,
    I had the same warning as @ ac67479 when I used the GATK3.7 version. If I run my job with 3.5 version, the job works well, without warnings.
    Here is the warning:
    WARN 14: 11: 27,231 Somatic Genotyping Engine - At Locus chr1: 7971781, we detected that variant context had alleles that not in PRALM. VC alleles = [A *, C], PRALM alleles = []

  • SheilaSheila Broad InstituteMember, Broadie, Moderator admin

    @jfpoloni
    Hi,

    Those WARNs are nothing to worry about. Have a look at my accepted answer above. I think the team will be changing those in the new version in GATK4.

    -Sheila

  • jfpolonijfpoloni BrazilMember

    Thank you so much @Sheila!!

  • bianxibianxi Member

    I run mutect2 on WGS in 366 segments with -L options, all others run successfully but 11:69774696-87688378 failed with this message,
    ERROR StatusLogger Unable to create class org.apache.logging.log4j.core.impl.Log4jContextFactory specified in jar:file:/usr/local/apps/GATK/3.8-0/GenomeAnalysisTK.jar!/META-INF/log4j-provider.properties
    ERROR StatusLogger Log4j2 could not find a logging implementation. Please add log4j-core to the classpath. Using SimpleLogger to log to the console...
    INFO 09:41:52,643 VectorLoglessPairHMM - Using OpenMP multi-threaded AVX-accelerated native PairHMM implementation
    [INFO] Available threads: 2
    [INFO] Requested threads: 1
    [INFO] Using 1 threads
    WARN 09:41:55,855 SomaticGenotypingEngine - At Locus chr11:69824156, we detected that variant context had alleles that not in PRALM. VC alleles = [A, G], PRALM alleles = []
    WARN 09:42:04,402 SomaticGenotypingEngine - At Locus chr11:69945734, we detected that variant context had alleles that not in PRALM. VC alleles = [C
    , G], PRALM alleles = []
    WARN 09:42:05,477 SomaticGenotypingEngine - At Locus chr11:69958659, we detected that variant context had alleles that not in PRALM. VC alleles = [C*, CT], PRALM alleles = []
    WARN 09:42:06,139 SomaticGenotypingEngine - At Locus chr11:69971956, we detected that variant context had alleles that not in PRALM. VC alleles = [A, AAAC], PRALM alleles = []
    WARN 09:42:06,139 SomaticGenotypingEngine - At Locus chr11:69971957, we detected that variant context had alleles that not in PRALM. VC alleles = [A, AAC
    ], PRALM alleles = []
    WARN 09:42:06,140 SomaticGenotypingEngine - At Locus chr11:69971958, we detected that variant context had alleles that not in PRALM. VC alleles = [A, AC], PRALM alleles = []
    WARN 09:42:07,488 SomaticGenotypingEngine - At Locus chr11:70003472, we detected that variant context had alleles that not in PRALM. VC alleles = [C, CA, CAAAAA, CAA
    , CAAA], PRALM alleles = []
    WARN 09:42:07,912 SomaticGenotypingEngine - At Locus chr11:70005641, we detected that variant context had alleles that not in PRALM. VC alleles = [A, G], PRALM alleles = []
    WARN 09:42:10,815 SomaticGenotypingEngine - At Locus chr11:70084818, we detected that variant context had alleles that not in PRALM. VC alleles = [T, TAA
    ], PRALM alleles = []
    INFO 09:42:18,099 ProgressMeter - 11:70287635 0.0 30.0 s 49.6 w 2.9% 17.5 m 17.0 m
    WARN 09:42:18,700 SomaticGenotypingEngine - At Locus chr11:70306737, we detected that variant context had alleles that not in PRALM. VC alleles = [T*, TGCGCGC], PRALM alleles = []
    WARN 09:42:31,427 SomaticGenotypingEngine - At Locus chr11:70575884, we detected that variant context had alleles that not in PRALM. VC alleles = [C, T], PRALM alleles = []
    WARN 09:42:31,872 SomaticGenotypingEngine - At Locus chr11:70582293, we detected that variant context had alleles that not in PRALM. VC alleles = [C
    , T], PRALM alleles = []
    INFO 09:42:48,100 ProgressMeter - 11:70867907 0.0 60.0 s 99.2 w 6.1% 16.4 m 15.4 m
    GATK finished (exit code -9)

    Here is the command line:
    MuTect2 -R /data/nextgen/ref/human_g1k_v37_decoy.fasta
    --cosmic /data/nextgen/ref/b37_cosmic_v54_120711.vcf
    --dbsnp /data/nextgen/ref/dbsnp_138.b37.vcf
    -L /data/nextgen/new_calling_intervals/variant_calling_intervals.agm.intervals
    -I:normal /data/nextgen/bqsr/Recal.ICGC_BM1_control_MB99_orginalRG.bam
    -I:tumor /data/nextgen/bqsr/Recal.ICGC_BM1_tumour_MB99_orginalRG.bam
    --normal_panel /data/nextgen/ref/merged_all.vcf
    -o /data/nextgen/bqsr/mutect2/sameAsMutect/cll/cll.agm
    --use_jdk_inflater
    --use_jdk_deflater

    Your help will be greatly appreciated!

  • shleeshlee CambridgeMember, Administrator, Broadie, Moderator admin

    Hi @bianxi,

    It looks like the run completed?

    You can ignore the ERROR StatusLogger Log4j2. Despite this error, the tool should run. Did you get any other error message for this run?

  • bianxibianxi Member

    It did not complete. There is no output. I did not get any other error message, either.

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie admin

    @bianxi Did you try running this with the latest release? I believe we're on 4.0.6.0. Testing that version will help us figure out if your issue might be version-specific.

Sign In or Register to comment.