Is this a valid allele in a vcf?

Question

[curious]

I have this variant given to me written as:

1:26697274:GCCA:CCCC

I think this is a "deletion and insertion" (is that a used term?) and this can be seen in the browser here

I don't think this would be validly written in VCF as:

CHROM POS REF ALT
1 26697274 GCCA CCCC

If I understand the VCF specification correctly we want to include the base before the variation as an achor, so if I wanted to harmonize the above 1:26697274:GCCA:CCCC to VCF spec, I would have to look that up the previous base with a software, which I think would translate to this:

CHROM POS REF ALT
1 26697273 CGCCA CCCCC

Is that right?

Answer 1

[Ram]

You don't need the anchor base for an indel like you do for deletions. The former 7274 entry is accurate.

[curious]

It looks like they do have an anchor for both insertion and deletion in the example 5.1.1

#CHROM POS ID REF ALT QUAL FILTER INFO
20 3 . C G . PASS DP=100
20 2 . TC T . PASS DP=100
20 2 . TC TCA . PASS DP=100

This is a really dumb question, but is my example and insertion-deletion, while the 5.1.1 example details one insertion and one deletion (both with anchors)

[Ram]

I guess you're right, they do have an anchor base for insertions as well, which in a way makes sense. More often than not, you see deletions where you see insertions so in the above example, callers would call the locus like so:

20 2 . TC T,TCA

In any case, go with what the specifications say. Use the additional anchor base.

[curious]

Thanks, but for this example would it be considered an insertion and deletion?

CHROM POS REF ALT
1 26697273 CGCCA CCCCC

Since it is a deletion of GCCA and insertion of CCCC

[Ram]

Yes, it would be considered an indel (which is a combination event where bases are both deleted and inserted). Since the number of bases in the end product is the same, I'd also call it an MNP but I'm not even remotely sure if I can do that so don't quote me on it.