[Fwd: FW: suggestions for 'POC' methods]

[Bruskiewich, Richard]

Hello everyone,

Michael is saying "everyone bake their own pie unless you decide to share
one pie" (i.e. collaborate). I was asking "are we cutting up one pie into
many pieces" and if so, how?

Does "a dedicated number space" mean "one pie cut into pieces" or "each
group has their own pie"? I've assumed the former, but even at that, the GO
has a number space and the PO/TO(*) has a number space. With Michael, I
wondered out loud whether or not the GO space was large enough to
accommodate the PO/TO too (7 digits == 10 million terms?). It would appear
not...

Richard

(*) PO/TO/PA or whatever. 'Traits' for Arabidopsis people only means
"phenotype" I suspect. For us crop people, the designation is a bit broader.
I still say that a plant is a plant is a plant, so all the ontologies for
traits, anatomy, development stages, etc. for plants belong under one number
space (which could be subdivided). The PO/TO split is highly artificial.
Note, however, that is not the same thing as saying that (mutant) phenotypes
won't involve some cross product with some more generic, orthogonal
phenotyping ontology. Maybe that is what Sue is thinking.

-----Original Message-----
From: Pankaj Jaiswal [mailto:pj37 at cornell.edu]
Sent: Saturday, March 23, 2002 3:46 AM
To: Sue Rhee
Cc: Michael Ashburner; R.BRUSKIEWICH at CGIAR.ORG; Leszek at missouri.edu;
CoeE at missouri.edu; polaccom at missouri.edu; curator at acoma.Stanford.EDU;
lreiser at acoma.Stanford.EDU; tberardi at acoma.Stanford.EDU;
steinl at cshl.org; (Susan Rutherford McCouch); ware at cshl.org;
gramene at brie.cshl.org
Subject: Re: [Fwd: FW: suggestions for 'POC' methods]


Hi,

I think, he is replying to the previous message from Richard Bruskiewich,
Leszek
should brief Richard about our decisions. I agree everybody is using a
dedicated
number space.

Pankaj


Sue Rhee wrote:
> 
> What message are you responding to? I don't think anyone is saying that we
> should not use dedicated numberspace.
> 
> On Fri, 22 Mar 2002, Michael Ashburner wrote:
> 
> >
> > I think I am missing something here.
> >
> > In my view each independent ontology should have its own id space.  The
> > syntax should be
> >
> > <prefix>:<zero-padded-integer>
> >
> > This means that a term, plus id, seen in isolation can immediately
> > be recognised with respect to its source ontology and there is no
> > risk at all of number clashes. GO uses a 7 fig digit; FB uses 9 fig
> > digits. DAGedit can handle both.
> >
> > If groups are collaborting on building/maintaining an ontology
> > (a la GO) then they divide their number space (see
go/numbers/go_numbers).
> >
> >
> > We then keep (eg in gobo) a register of prefixes.
> >
> > What is wrong with this ?
> >
> > Michael
> >
> 
>
----------------------------------------------------------------------------
-
> Sue Rhee                                rhee at acoma.stanford.edu
> The Arabidopsis Information Resource    URL: www.arabidopsis.org
> Carnegie Institution of Washington      FAX: +1-650-325-6857
> Department of Plant Biology             Tel: +1-650-325-1521 ext. 251
> 260 Panama St.
> Stanford, CA 94305
> U.S.A.
>
----------------------------------------------------------------------------
-

-- 

******************************************
Pankaj Jaiswal, Ph.D.                                   
Postdoctoral Associate
Dept. of Plant Breeding                             
Cornell University                                   
Ithaca, NY-14853, USA   

Tel:+1-607-255-3103 / Fax:+1-607-255-6683
E mail: pj37 at cornell.edu
http://www.gramene.org   
******************************************