VCF to genotype file of specific format

Question

[David.shaw]

I have a VCF per individual.

I'd like to take it, and convert it to the following format so that I have the alleles of the individual in the last two columns:

rsid chr pos ref alt

rs102 chr1 34 A G

Any package that does this already?

Answer 1

[Shane McCarthy]

For REF and ALT in the last two columns:

bcftools query -f '%ID %CHROM %POS %REF %ALT\n' input.vcf

For the actual alleles of the individual in the last two columns:

bcftools query -f '%ID %CHROM %POS[ %TGT]\n' input.vcf | tr "/" " "

[David.shaw]

This doesn't quite work in regards to the actual alleles if some are phased and some aren't (because they are rare).

This worked:

bcftools query -f '%ID %CHROM %POS[ %TGT]\n' ../131.vcf | tr "|" " " | tr "/" " " | less -S

[Shane McCarthy]

Then replace the tr command at the end with

tr "/|" " "

Answer 2

[Vivek]

awk '! /\#/' file.vcf | awk -F '\t' '{print $3"\t"$1"\t"$2"\t"$4"\t"$5}'

[David.shaw]

No, This doesn't handle more than biallelic SNPs.

Answer 3

[Pierre Lindenbaum]

GATK VariantsToTable https://www.broadinstitute.org/gatk/blog?id=7089

or my tool BioAlcidae see Taking genotypes out of a vcf file