Can funcotator remove mutation data from dbsnp

Follow

Luo Hao

• January 02, 2024 02:11

REQUIRED for all errors and issues:
a) GATK version used:

gatk-4.4.0.0
b) Exact command used:  

gatk Funcotator \
-R $ref/Homo_sapiens_assembly38.fasta \
-V ../${array[1]}.filter.vcf \
-O ${array[1]}.maf \
--output-file-format MAF \
--data-sources-path Funcotator_data/funcotator_dataSources.v1.7.20200521s/ \
--ref-version hg38 \
--annotation-default tumor_barcode:${array[1]} \
--annotation-default normal_barcode:${array[0]} 
done

c) Entire program log:
Hi, I use funcotator to call somatic mutations as mentioned above. Can funcotator remove the mutation data in dbsnp by default? or I need extra removal? Is there a parameter to remove these mutations?
Thanks!

0

• 

  David Roazen

  • January 12, 2024 21:48

  Hi Luo Hao,

  If you want to remove known germline variants from your callset, you can either do this while running a somatic caller such as Mutect2, or post-calling using a tool like SelectVariants. Funcotator itself cannot do this, as it assumes that you've already filtered your callset using one of these other tools.

  Regards,

  David

  0

  Comment actions Permalink
• 

  Luo Hao

  • January 17, 2024 03:49
  • Edited

  Hi David,

  I run Mutect2 before running Funcotator as follows:

  gatk Mutect2 \
  -R $ref/Homo_sapiens_assembly38.fasta \
  -I ${array[0]}.sort.dedup.baserecal.bam \
  --tumor-sample ${array[0]} \
  --germline-resource af-only-gnomad.hg38.vcf.gz \
  --af-of-alleles-not-in-resource 0.0000025 \
  --panel-of-normals 1000g_pon.hg38.vcf.gz \
  --output ${array[0]}.vcf
  
  gatk FilterMutectCalls \
  -R $ref/Homo_sapiens_assembly38.fasta \
  -V ${array[0]}.vcf \
  -O ${array[0]}.filter.vcf

   

  I am not sure could the "--germline-resource af-only-gnomad.hg38.vcf.gz" parameter remove the dbsnp site or I need to do something extra?

  Thanks!

   

   

  0

  Comment actions Permalink
• 

  Can Kockan

  • January 30, 2024 18:19
  • Edited

  Hi Luo Hao , if your intention was to remove germline variants as David previously mentioned, the gnomAD germline resource should work fine. It should even be a more thorough resource than dbSNP in terms of the number of samples, number of variants, better allele frequencies, etc. so you shouldn't need to do anything extra. Of course, if you do see a variant that you believe shouldn't be there, let us know.

  0

  Comment actions Permalink

Please sign in to leave a comment.