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

[Pankaj Jaiswal]

Dear Sue,

I agree, we need to divide the number space and Leszek will follow up on the
number space limits for respective groups. This way we will definitely avoid
mix-ups. The reason I asked for a PO (preferred) identifier as compared to PA,
is that , we want to bring in the developmental stages also under one single
umbrella of PO. While doing this I intend to use the higher level terms curated
by TAIR, if it confines to developmental stages in rice/other crops and curating
the terms which may be either specific or generic to other monocots. This is one
aspect . Another one is that at present I have the traits represented in TO as
very specific to rice and I am still working on them. As soon as they develop
into a more generic format representing more plants, I intend to make them as
"part of" the PO. Therefore I believe PO will serve the purpose and can easily
serve the purpose in a long run by providing an umbrella coverage to plant
-anatomy, -development + developmental stages and -traits. Please do not view it
in the context that why am I keeping the TO's separate, its all that I do not
want them to acquire the space on PO until they are mature enough to represent a
wide array of traits.

On the other hand the way our Ontology database works is that the primary
identifiers are quite often very unstable and they do get generated on the fly
while updating the database. The accessions such as the GO and TO IDs, are all
unique in the sense that they are recognized by the ontology types
"component/function/process" and "trait". This way the type of ontology
determines what should be the prefix GO/TO. presently we have term type as
"anatomy" for the terms you see from TAIR. We know its possible to recognize the
different identifier such as TAIR/MAIZEDB/etc. though it would benefit in the
long term if we call it by the same identifier such as PO for "anatomy and
development". We will try to separate the term types under PO as "anatomy and
development" and later on at some point "trait" will also join, but that will
depend on how things shape up while I curate the trait terms and the decisions
taken up by GO to represent the traits either as a single ontology term or as a
cross product. I am sure it will go the later way.

On an all as we progress we will mature and gain from experiences of GO and
certainly of your and TAIR's to help make the POC more distinct and meaningful
to the plant community. 

As you said I will modify the search queries for GO/PO/TO in the browser. One
will be able to select the ontology type for search.

Lastly but not the least, since we are designating the number space to every
group, I am not hesitant in using the TAIR IDs as either DBXref/synonyms. Next
time when we update the DB I will provide the PO ID for the anatomy/development
term, and will use the TAIR ID as either the synonym or DBXref, if you agree.
Could you please tell me if you have already incorporated these terms/IDs in
your curation or still in the process of using them. It will help us at Gramene
in setting up the tables and browser. Even though I am suggesting all these view
points you will agree that it takes a certain amount of resources and time and
everyone of us is limited within the scope and mandate, we will try our best.

Thanks

Pankaj

Sue Rhee wrote:
> 
> Hi, Pankaj,
> 
> I don't see that Michael is suggesting one versus the other in what I can
> make out of the message below. The important issue about the identifier is
> that we divide up the numberspaces among the three groups such that we do
> not use the same number (accession). We reached an agreement on this
> yesterday, yes?
> 
> We have been designating a seven digit number and if this is suitable for
> everyone, each group can take 10,000 numbers to start with (We've used up
> ~200 in Anatomy).
> 
> In terms of the prefix, the reason I see for using a prefix to this number
> accession is to distinguish the shared plant anatomy ontology from
> different ontologies (e.g. GO, and inevitabley more ontologies to follow).
> Since we agreed on making a shared plant anatomy ontology, using PA as the
> prefix sounds ok to me so long as it is limited to 'Plant Anatomy
> Ontology' and that we stick to the numberspaces for the numbers (so that
> we don't create yet another accession for anatomy for Arabidopsis, for
> example).
> 
> PO sounds a bit too general for this ontology. If we agree on making a
> combined plant ontology for Developmental Stages (or temporal), we can
> perhaps designate a different prefix. At this point, it is not clear to me
> that we can develop a combined temporal ontology for plants, but it is
> certainly worth a try if you are willing/interested. It is likely that the
> development of Trait and other types of ontologies would occur as a
> collaboration of more than plant groups, and may end up with a differnt
> prefix.
> 
> Sue
> 
> On Wed, 20 Mar 2002, Pankaj Jaiswal wrote:
> 
> > Dear Everyone,
> >
> > I am forwarding you an e mail from Michael, Some of you must have seen it
> > earlier also. There he has tried to resolve the issue of identifiers (agenda no.
> > 2 from yesterday's conference call) and suggested to use PA "Plant Anatomy"
> > which is a more generic identifier. I am strongly seconding his suggestion and
> > would recommend using "PO" Plant Ontology instead of the PA or the identifiers
> > from respective databases. I would invite everyone to discuss pros and cons of
> > it while building up the database. It is infact has turned out to be a major
> > issue and should be resolved ASAP.
> >
> > Thanks
> >
> > Pankaj
> >
> >
> > "Michael Ashburner (Genetics)" wrote:
> > >
> > > Curious what emails cross one's desk !
> > >
> > > If I may express a view, it for option 1. BUT John Richter
> > > should be consulted how DAGedit will cope with mixed if prefixes.
> > > The thing is for option 1 is that should any one want an ontology
> > > just for Arabidopsis then the Zea etc specific terms could be stripped
> > > in a thrice with an editor.
> > >
> > > The other way is to have a generic prefix for these terms, eg. PA
> > > for "Plant Anatomoy" and either use these in your individual databases
> > > or have your own id's as synonyms.
> > >
> > > Two points, one minor.
> > > Not
> > >              adult leaf sensu Arabidopsis thaliana TAIR:
> > > but
> > >              adult leaf (sensu Arabidopsis thaliana) TAIR:
> > >
> > > Two, I am no botanist but could these not be made a bit broader
> > > eg
> > > (sensu Brassicae)
> > > (sensu Graminae)
> > >
> > > (???or sensu dicots, senu monocots) and then only go to a more
> > > specific sense if the biology so demands ?
> > >
> > > Hope I am not butting in when inappropriate.
> > >
> > > Michael
> >
> > --
> >
> > ******************************************
> > 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
> > ******************************************
> >
> 
> -----------------------------------------------------------------------------
> 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   
******************************************