To celebrate the release of GATK 4.0, we are giving away free credits for running the GATK4 Best Practices pipelines in FireCloud, our secure online analysis portal. It’s first come first serve, so sign up now to claim your free credits worth $250. Sponsored by Google Cloud. Learn more at

Call "KEEP" "REJECT" option to fine tune filters for poor_mapping_region_alternate_allele_mapq fail

dudududu Member

I am using the the MuTech program in standard conditions (with --enable_extended_output) with a matched Normal/tumor couple with the
following code at TP53 locus. We have confirmed the presence of a mutation in this tumor by sanger sequencing at chr17 position 7578526.

I used the following code :
java -Xmx2g -jar muTect-1.1.1.jar --analysis_type MuTect --reference_sequence fusion2.fasta --enable_extended_output --dbsnp dbsnp_132_b37.leftAligned.vcf --cosmic b37_cosmic_v54_120711.vcf --intervals chr17:7571720-7590868 --input_file:normal bloodsorted.sorted.bam --input_file:tumor Tumor02878.sorted.bam --out result.call_stats.txt --coverage_file example.coverage.wig.txt

Using MuTech, mutation at this position get the judgement 'REJECT'. I looked in the extended output and found the the reason of failure
was : poor_mapping_region_alternate_allele_mapq.

I am puzzle, any suggestions ? (I attach file result result.call_stats.txt)

I would like to decrease the stringency of the coresponding filter to get the judgement 'KEEP' for the mutation that was confirmed. I was unable to find which option to use in the program and with which value ( I tried --required_maximum_alt_allele_mapping_quality_score option with different values but it did not change 'KEEP", "REJECT" status).

Hope I was clear in my explanations, I am new in the forum and not used to post question.


  • kcibulkcibul Cambridge, MAMember, Broadie, Dev

    Hi -- this is an interesting case, what sort of sequencing technology is this?

    The problem is that the maximum mapping quality score for the alternate is actually a negative number (see column t_ref_max_mapq). I'm not sure how this could be if this is a valid BAM file. Would it be possible for you to share a small BAM file containing just reads intersecting this locus so I can reproduce the problem. It's possible the BAM is fine and it's a bug in the caller somehow.

    Alternatively, can you use another tool like Samtools to view the reads intersecting this locus and tell me what the mapping quality scores are? They should not be negative of course.

  • ailisailis Member

    Hi Kcibul and Stan,

    Were you able to resolve this?

    I've tried running MuTect using the default HC parameters and all of the mutations have "poor_mapping_region_alternate_allele_mapq" in the "failure_reasons" column. I thought it was because I was being too ambitious by running it on RNA-seq data. But having read this post, perhaps my BAM file isn't in the correct format or there's something else going on?


  • kcibulkcibul Cambridge, MAMember, Broadie, Dev

    I didn't hear back from Stan about the negative value for t_ref_max_mapq, but what sort of values are you seeing for that field in the mutations that are only rejected due to "poor_mapping_region_alternate_allele_mapq"?

  • hns04hns04 Member

    Hi Kristian,
    I am attaching the file which have the t_ref_max_mapq values, they range from -1 to 3. Also, my samples are RNA seq samples for mouse models. I get ~600k calls and all of them are reject calls. I am attaching a small file with 500 calls for you to take a look at.

  • hns04hns04 Member

    Hi, Anyone else facing the same problem or know the solution to it?

  • hns04hns04 Member

    Hi can any of the developers help us with this issue?.

  • GourabSahaGourabSaha IndiaMember

    Hey i am having the same problem all my calls are 'REJECT' ed . What to do? Please someone reply

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie

    @GourabSaha Can you describe in detail what is your experiment, datatype and what is the rejection reason listed in most calls in the callstats file?

  • GourabSahaGourabSaha IndiaMember

    fstar_tumor_lod,possible_contamination,normal_lod,poor_mapping_region_alternate_allele_mapq ... These are reasons shwowing in 99% cases.. my experiemnts is based on 400 genes exons custom sequencing of tumor normal pair in Hiseq2500. i did indel realignment , base quality recalibrator of bam file before running mutect. what to do now?

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie

    It sounds like you have some data quality issues. Have you looked at the data in the region around some of the sites? How much sequence depth do you have? And can you post the console output showing the filtering summary?

  • mceredamcereda ItalyMember

    Hi All,
    I have the same problem of negative value in the t_alt_max_mapq (e.g. -106) and my variants are REJECTed as "poor_mapping_region_alternate_allele_mapq". Do you have solution for this issue?

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie

    You can try the brand new version of MuTect, called MuTect2, which is now available directly in GATK (3.5). It also calls indels, not just SNPs.

  • perryeperrye New Haven, CTMember

    I am currently encountering this problem, with mutations being rejected based on poor_mapping_region_alternate_allele_mapq fail. These sites have negative values listed in the Mutect output, even though the mapq score for these reads is high when I view the .bam file with IGV. Has a source of this problem been identified? Has anyone confirmed that the issue is resolved in MuTect2? Thank you!

Sign In or Register to comment.