[Open-bio-l] BioSQL schema: some questions

[Hilmar Lapp]

> -----Original Message-----
> From: Ewan Birney [mailto:birney@ebi.ac.uk]
> Sent: Sunday, April 28, 2002 2:51 PM
> To: Hilmar Lapp
> Cc: Chris Mungall; OBDA BioSQL (E-mail);
> gmod-devel@lists.sourceforge.net
> Subject: RE: [Open-bio-l] BioSQL schema: some questions
> 
> 
> On Sat, 27 Apr 2002, Hilmar Lapp wrote:
> 
> > > 
> > > molecule = where it came from (eg mRNA)
> > > 
> > > alphabet = how it is encoded (DNA/RNA etc)
> > >
> > 
> > Shouldn't then molecule be an attribute of Bioentry? 
> Likewise, shouldn't alphabet be added to Bioentry?
> 
> One would hope these are both biosequence properties, in partuclar the
> alphabet
> 

Well, my original argument was that that would mean that a bioentry could have many biosequences, with those attributes being different from one to another (right now, the relationship is identifying, i.e., there can be only one biosequence for a particular bioentry). Or, those attributes would possibly change their value from one biosequence version to another. I don't really know whether that can possibly be the case.

> 
> Swissprot has alot of DR lines with alot of databases, most 
> of which one
> does not want to instatiate a database entry for (although one
> could...) each which has it own pecularities about id tracking.

What would be the downside of instantiating a database entry for each of those? Wouldn't you have to match the dbname against biodatabase.name anyway if you want to look up the referenced entry? I guess only if you wanted to do really that, but not if it's just for record-keeping aimed at round-tripping, right?

It also strikes me that you need only dbname and accession in Dbxref to uniquely reference a record, whereas you need (at least) the version, too, in bioentry.

Related to that, why is there only accession and version (you see, dbxref is the only one without a version) in Remote_seqfeature_name? Because the dbname is implicitly the database the bioentry is linked to the seqfeature is linked to?

	-hilmar
-- 
-------------------------------------------------------------
Hilmar Lapp                            email: lapp@gnf.org
GNF, San Diego, Ca. 92121              phone: +1-858-812-1757
-------------------------------------------------------------