Somatic variant calling mutect2 tumour only

Question

[cellulebioinfobiscem]

Hi,

I have to call somatic variant for mutate sample with no matching normal.

I have read the GATK mutect2 page, but still not understand the tumor only mode..

They call that tumor only mode and say "call on each normal sample in this mode", but I don't have normal, so I don't understand what I should do.

Thanks for any explanation:

"

(ii) Tumor-only mode This mode runs on a single type of sample, e.g. the tumor or the normal. To create a PoN, call on each normal sample in this mode, then use CreateSomaticPanelOfNormals to generate the PoN.

  gatk Mutect2 \
   -R reference.fa \
   -I sample.bam \
   -O single_sample.vcf.gz

To call mutations on a tumor sample, call in this mode using a PoN and germline resource. After FilterMutectCalls filtering, consider additional filtering by functional significance with Funcotator.

  gatk Mutect2 \
  -R reference.fa \
  -I sample.bam \
  --germline-resource af-only-gnomad.vcf.gz \
  --panel-of-normals pon.vcf.gz \
  -O single_sample.vcf.gz

"

Answer 1

[benformatics]

Your question is confusing... (ii) clearly provides you a command to run (-I can be any sample).

  gatk Mutect2 \
   -R reference.fa \
   -I sample.bam \
   -O single_sample.vcf.gz

The reason there are no explanations on how to do an analysis without a normal/control sample is because frankly you should never do this. They (the GATK team) are not going to provide you with instructions on doing something that they don't think will provide strong results. Nonetheless, you can still run the command as provided.

Relevant GATK forum posts:

• https://gatkforums.broadinstitute.org/gatk/discussion/2814/mutect-analysis-without-a-paired-tumour-normal-analysis-criteria
• https://gatkforums.broadinstitute.org/gatk/discussion/6840/can-mutect2-be-applied-in-tumor-only-mode-for-somatic-variant-detection
• https://gatkforums.broadinstitute.org/gatk/discussion/12091/calling-somatic-variants-without-matched-normals-using-gatk

Answer 2

[newbio17]

As benformatics explained, tumor-only variant calling is not recommended and should be avoided if possible.

With that said, GDC sometimes receive only tumor samples and have prepared PON, which is available on their page (scroll to bottom), to process the samples upon submitter's request: GDC reference files (Requires dbGaP access)

You may also want to visit gnomAD to get the germline resource for Mutect2 workflow.

There are other discussions on Biostars such as Question: tumor only variant calling tools which discusses other available tools as well.

Hope this helps.

[cellulebioinfobiscem]

Thanks for your answer but I use BALBcJ and C57BL6NJ specific reference. I don't find any germline resource for this two mouse strain.

[benformatics]

Then try using reference data from the mouse genome project... it is also available as very large BAM files.

ftp://ftp-mouse.sanger.ac.uk/current_bams

Furthermore it still seems like you have not understood my initial answer... all you have is a single sample. Why not provide the reference as the genome for the specific mouse strain you are attempting to investigate?

Or a somewhat painful way to do this would be to call your mutations with the standard mouse genome (which I just looked up is C57BL/6J).

Then take the resulting SNPs (i.e. VCF file) and remove from those lists the known variants that differ between the reference strain and the strain you are analyzing (https://www.broadinstitute.org/files/shared/mammals/mouse/snp/gapmapselect.html). This should give you somatic variants.

Frankly I would suggest you find somebody or a paper where they did a similar analysis to what you want to do any copy their methods.