Easy Way To Change Reference Alleles In Large Vcf File?

Question

[Kking]

Before I begin coding, does a tool already exist that allows you to easily switch reference alleles in a large VCF file (~400K variants) based on a reference genome, re-encoding all the genotypes properly?

We have a large amount of legacy data in PLINK format that we would like to use with some of the modules in GATK's VariantEval method to compare with whole exome data. I tried converting the PLINK data to VCF using PLINK v1.08. However, it does not have a mechanism for specifying the reference allele, and the output did not match our sequencing files.

Answer 1

[Baojian Fan]

I have the same problem to convert PLINK files to VCF files. I loaded the PLINK files into a project using PLINK/SEQ and then output them to VCF files using the command write-vcf. However, the resulting VCF files have mismatched alleles with REFDB. Does anyone know how to solve this problem? Thanks.

[lh3]

This is not an answer. Please add this as a comment to the original question. Thank you.

Answer 2

[Raony Guimaraes]

May be you could do something like this:

grep "^#' yourFile.vcf > newvcffile.vcf

grep -v "^#" yourFile.vcf | awk '{printf $1, $3, $2, $4... }' >> newvcffile.vcf

[Hypotheses]

Can you help explain more what you mean by ^# in grep? how does this help changing the reference allele coding? Thank you.