[Open-bio-l] RE: seqfeature

[Thomas Down]

On Tue, Apr 30, 2002 at 05:49:08PM -0700, Hilmar Lapp wrote:
> Seqfeature_location does not have a frame attribute (nor a phase). On purpose?

Phase (at least in the way it's normally used) is quite specific
to coding exons.

I'd actually argue that no `biological' meaning should ever
be attached to locations (let alone /parts/ of a location).
If it's important to you, create individual `exon' features
and put the phase as a qualifier_value on that.  (Note: this
is coming from the BioJava `location-is-a-set-of-integers'
point of view.  BioJava locations don't even have strand
information, so they're already a little different from the
BioSQL model).

If you really, really, want phase on individual location spans,
why not put it in the location_qualifier table.  That at least
makes it clear that phase isn't a property that every location
is expected to have.


     Thomas.