Splicing terms (e.g. SO:0001629) annotation for variants with multiple representations #674
Description
I am wondering how to correctly annotate a variant that can be shifted over an intron/exon-boundary, and thus has representations both outside and inside the bounds defined by the corresponding SO terms: is it sufficient that there is at least one variant representation fulfilling the SO term definition, or does it need to hold for all representations?
Are there any general recommendations on this from the SO team on how to deal with variant shifting, or is this left up to annotation tools like SnpEff or VEP?
Maybe the definitions of specific SO terms could clarify this issue, to make this less ambiguous? For example, the definition of SO:0001574 could read:
A splice variant where at least one representation changes the 2 base region at the 3' end of an intron.
(If possible, documenting a general approach to this problem might be nicer, to keep the definitions short.)