GO and Cell ontologies

[Pankaj Jaiswal]

Sorry for the late reply, I agree with Sue, but there is another problem.
If we remove the cell types from PO or for that reason in any anatomy 
ontology, then there is no mechanism (at least in the present day) we 
can connect to say X-cell is a part_of Y-Tissue/organ from 
plant/fly/any_organism.

As an alternative, I still strongly believe, that all the cell types be 
maintained in individual organism/clade specific anatomy ontologies and 
whenever required, a complete but flat vocabulary of cell type (aka cell 
ontology) be created.  This will also help to maintain the sensu 
aspects. However we may lose the comparative aspects, the moment we go 
down the sensu path.

-Pankaj


Sue Rhee wrote:

> I can think of two pragmatic issues at hand with respect to the
> possibility of obsoleting cell type terms in PO. One is whether the scope
> of the cell type node within PO is comparable to that of the plant cell
> type node within CL. The other is how quickly the databases using PO can
> switch to using CL for the cell types. Before these two issues are
> addressed, we cannot obsolete the cell type node within PO.
> 
> Regarding the first issue, PO's current goal is to accomodate plants
> within the angiosperm taxa only. This means that we may not cover cell
> types in lower plants that might be useful to have in a more generic cell
> type onology. On the other hand, even within this subdomain of plant
> taxanomy, we could have some quite granular terms (e.g. that are
> applicable to only a few species) that might be too specific for the scope
> of CL. I'm amenable to changing the scope of either or both of the
> ontologies to allow the mutual exclusivity of the two ontologies in terms
> of primary terms, but this is something that needs to be agreed upon all
> pertinent developers of both of the ontologies.
> 
> Regarding the second issue, it took TAIR 8 months to switch from TAIR
> anatomy to PO structure ontology, with plenty of time to plan for it. So
> this is probably something we need to plan quite a bit in advance..
> 
> Sue
> 
> 
> On Mon, 25 Apr 2005, Pankaj Jaiswal wrote:
> 
> 
>>
>>Chris Mungall wrote:
>>
>>
>>
>>>Regarding PO and CL - at the plant ontology meeting I suggested simply
>>>removing all cell terms from PO - is there any need to duplicate them?
>>>
>>
>>
>>I agree on maintaining the cell types in cell.ontology, but the BIG
>>questions are:
>>	#1 How many different ontologies are we expecting a database to use in
>>their annotations?
>>	#2 What is the border line for each ontology, so that we do not overlap?
>>	#3 Is it good for the users or is it confusing? Users being both the
>>actual wetlab scientists and the curators.
>>	# Is there a better way of semantically using an integrated version of
>>all the required ontologies, despite being maintained independently?
>>
>>-Pankaj
>>