Regular Expression help for (bcf/vcf)tools filter manipulation

Question

[morovatunc]

Hi,

I annotated the flanking regions of SNV mutations with vcftools fill-fs tool. In example;

20  10639675    .   C   G   .   .   Callers=broad,dkfz,muse,sanger;NumCallers=4;VAF=0.3571;repeat_masker=L1M5;t_alt_count=25;t_ref_count=45;Variant_Classification=Intron;FS=t[C/G]c

it added FS=t[C/G]c where;

• T is the 5prime flanking base and C is the 3 prime flanking base.
• C is the reference and G is the alternated base.

I would like to subset mutations which are similar to followng pattern.

N[T/A]G or C[A/T]N

If you can answer my question by using bcftools/vcftools, it would be very good but I am okay with any other solutions. I know bcftools view -i has a filter option but I couldnt figure it out.

Thank you very much for your help,

Best regards,

Tunc.

Answer 1

[Petr Ponomarenko]

grep -E '\[T/A]g|c\[A/T]|^#' input.vcf > output.vcf

-E is for regexp grep

[ is to search for [ in regexp instead of opening construct []

| is for OR

^ is to search for the first symbol so ^# is to preserve the header of vcf

in place of \[T/A]g you may want to use FS=?\[T/A]g if INFO field has such constructs for data other than FS

[morovatunc]

Thank you very much.