Best way to get Allele Frequency in dbSNP file of SNPs in a different VCF file?

Question

[markgodek]

Hi,

I have some VCFs generated by Mutect 2. For each variant in these files, I want to get the allele frequency from dbSNP and 1k Genome VCFs I have.

Could you recommend a tool to do this?

I considered writing a python script to do it, but thought there is a better way than iterating over both dbSNP and 1k Genomes for every line in my Mutect 2 output.

Thanks.

[brunobsouzaa]

Why don't you use Ensembl VEP to annotate your VCFs?

[markgodek]

We decided to use GATK best practices so we're using their Funcotator for functional annotation, but the "Frequency data for co-located variants" function of VEP does look promising .Thanks.

[Ram]

Take a look at bcftools annotate. You may also want to check if your input is normalized (left aligned, parsimonious and multi-allelics split) before using bcftools annotate. You can do the pre-processing steps using bcftools norm or vt decompose + vt normalize

[markgodek]

Thanks. I'm new to all this so when vcftools and bcftools had given me errors about multiallelic sites, I just removed them with SelectVariants --restrict-alleles-to BIALLELIC

So after pre-processing with bcftools norm, I should be able to do something like this?

bcftools annotate -a 1000G_phase1.snps.high_confidence.b37.vcf.gz -h annotations.hdr -c CHROM,POS,ID,INFO/1000G_AF:=INFO/AF myVCF.vcf

with annotations.hdr being something like

##INFO=<ID=1000G_AF,Number=.,Type=Float,Description="Allele Frequency from 1000K genomes high confidence SNPs">

[Ram]

Pass a 100 variants through your bcftools annotate and verify a few - that way, you will know it works. You may want to use the --collapse parameter to make sure comparisons take CHROM, POS, REF and ALT into account to match the 2 VCFs.

[markgodek]

Thanks for your help. I was using bcftools annotate to change chromosome names in an earlier step, but didn't know it also had this function.

I've taking a deeper dive into bcftools and vcftools and it's really making my project easier.

Thanks again.