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

Allele depth difference between mutect VCF and IGV

coonyacoonya Member
edited April 2015 in MuTect v1

Dear Mutect team,

I usually use Mutect for calling somatic mutations. I faced problem with depth in variant position.

We use both paired (tumor- matched normal) and tumor only analysis. For instance, Mutect called somatic mutations as follows:

7 55259445 . C T . PASS SOMATIC;VT=SNP GT:AD:BQ:DP:FA:SS 0/1:62,9:24:71:0.127:2 0:0,0:.:0:0.00:0
7 55259446 . A T . PASS SOMATIC;VT=SNP GT:AD:BQ:DP:FA:SS 0/1:63,7:26:70:0.100:2 0:0,0:.:0:0.00:0

These two positions have approximately 70x for total reads and 10% variant allele frequencies in VCF file

However, the total depth and variant allele depth in those position were 200x and 100x (50%) in IGV.

Please let me know why VCF present quite less read depth than IGV. Are there any reads filtering steps for calling variants?

I have one more question. If there are overlapped reads that are originated from insert size ( e.g) 250bp - insert size, 150 bp * 2 - sequencing read), how does Mutect treat the reads?



  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie

    Hi @coonya,

    Yes, there are some filtering steps before calling variants that could explain why you see this difference.

    I don't understand your second question, can you please clarify?

  • Hi @Geraldine_VdAuwera,
    I appreciate your comment. If possible, please let me know what what filtering steps to be able to affect read depth does Mutect use.

    If I generate average insert size is 250 and sequence those reads with 150bp * 2. A middle of those reads (approximately 50bp are overlapped) will sequence twice. If Mutect detect somatic mutation in that position (overlapped regions), how does Mutect count variant allele? Can it be one or two?

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie

    @coonya, MuTect applies filters that remove reads that do not pass mapping quality thresholds, duplicate reads, and other similar reads that are not usable. You can see the read filtering summary in the log. If you're comparing to what you see in IGV, be sure to set your IGV preferences to not display these types of reads, or take them into account when you count them.

    MuTect counts each read as a separate observation regardless of pairing information.

Sign In or Register to comment.