This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Remove ambiguous calls in the VCF file

Hi everyone,

I have a VCF file with multiple ambiguous ref/alt calls at some positions of the genome with ref allele type Y, R, M, K, S, W (i.e. two-base ambiguity codes). e.g.

3       60830534        .       M       C       101     .       .       GT:DP:A:C:G:T:PP:GQ     1/1:24:0,0:14,9:0,0:1,0:1038,0,808,782,114,898,883,114,101,806:101,

is there any way to remove them all from the VCF file?

Kind Regards sohail

vcf
• 4,032 views • link
updated by ATpoint ATpoint • written by SOHAIL SOHAIL

1 answer

This will print the header of the VCF and only those entries where both REF and ALT are A/C/T/G:

awk '$1 ~ /^#/ {print $0;next} {if ($4 ~ /A|C|T|G/ && $5 ~ /A|C|T|G/) print $0}' in.vcf > filtered.vcf
• 1 views • link
written by ATpoint ATpoint

@ATpoint Thanks for the reply!

Your command is working good for the type of VCF file where only variants are only called (i.e. at both columns of REF/ALT A/T/G/C should be present).

However, my VCF file is called with all genotypes of the genome "all-positions" (either homo ref or homo alt or het sites) together with ambiguous variant call set. and column 5 (ALT) of VCF might be filled with the period (i.e. dot symbol) e.g.

3       60830534        .       M       C       101     .       .       GT:DP:A:C:G:T:PP:GQ     1/1:24:0,0:14,9:0,0:1,0:1038,0,808,782,114,898,883,114,101,806:101,
3       60830535        .       C       .       101     .       .       GT:DP:A:C:G:T:PP:GQ     1/1:24:0,0:14,9:0,0:1,0:1038,0,808,782,114,898,883,114,101,806:101,

When i modified the command with following, ambiguous call is still there 

awk '$1 ~ /^#/ {print $0;next} {if ($4 ~ /A|C|T|G/ && $5 ~ /.|A|C|T|G/) print $0}' in.vcf > filtered.vcf

am I doing any mistake?

• 1 views • link
updated by finswimmer finswimmer • written by SOHAIL SOHAIL

Sorry, I do not get it. From the two lines above, the one where REF is M and the one with ALT ., which of these should be removed?

• 1 views • link
written by ATpoint ATpoint

@ATpoint, The lines with M (and others Y, R, W, K, S, (i.e. two-base ambiguity codes) ) in VCF file will be removed.

edit: any help???

• 1 views • link
written by SOHAIL SOHAIL

Log in to answer this question.

More posts like this