I am writing a small script to write a VCF file from short multiple sequence alignments, and stumbled into an interesting corner case.
This is the way a deletion should be represented in VCF format (I assume version 4+):
NC_000016.9 2138199 rs1799758 GTGAG G . . RS=1799758;...
But what happens if the deletion is at the first position? Immagine the following two aligned sequences:
>reference
AGTC
>sample
-GTC
How should it be encoded in VCF format? According to this thread on SEQAnswers, it might be necessary to include the first position after the end of the deletion, like this:
chromosome 1 AG G . . [...]
Is this a VCF 4+ correct way to encode such variants?