Moving AF tag to new column in a VCF file

Question

[Christian]

Hi,

I have called some somatic variants in tumor-only mode with Mutect2. The first line of the output VCF file looks like this:

#CHROM POS   ID   REF   ALT   QUAL  FILTER INFO  FORMAT 156390
chr1  17961  .    TG   T    .    PASS  .    GT:AD:AF:DP:F1R2:F2R1:SB    0/1:178,59:0.251:237:96,32:72,24:37,141,10,49

Where "156390" is the sample ID. For a downstream tool, I would need the variant allelic fraction (AF) to be either in column 8 (INFO) or column 11, formatted as AF=x.

I have tried to solve this with awk:

awk -F '\t' 'BEGIN{OFS="\t"} {split($10, arr, ":"); $11="AF="arr[3]; print}' test_156390.vcf > test.vcf

But this is clearly a suboptimal solution - the new column 11 looks right, but AF= is now also the column name, and is also appended to every line in the header.

Is there any better way of doing this?

Thank you

Christian

Answer 1

[Pierre Lindenbaum]

extract the vcf as a tab delimited file, something like (not tested)

bcftools query -f '%CHROM\t%POS\t%REF\t%ALT[\t%FORMAT/AF]\n' in.vcf > out.tsv
bgzip out.tsv
tabix <something> out.tsv.gz

then use bcftools annotate -a out.tsv.gz <something...> in.vcf to insert back a new INFO field in the vcf

[Christian]

Thank you for pointing me in the right direction! This is how I solved it and applied it to every VCF in dir. I could not quite get bcftools annotate to work, likely due to a local server issue, but vcf-annotate works fine as well.

for i in *.vcf; do echo "annotating $i"; 
bcftools query -f '%CHROM\t%POS\t%REF\t%ALT\t[%AF]\n' $i > tmp.tsv
bgzip -f tmp.tsv
tabix -f -b 2 -s 1 -e 2 tmp.tsv.gz 
cat $i | vcf-annotate -a tmp.tsv.gz -c CHROM,POS,REF,ALT,INFO/AF -d key=INFO,ID=AF,Number=1,Type=Float,Description='Variant allelic fraction' > "${i//.vcf}"_vaf.vcf
done