progress and thought questions

[Vincent, Leszek]

Toby, Pankaj & colleagues,

I offer some brief thoughts in response to the matters raised in the
dialogue below. I've inserted these below, prefixed by XXX. - Leszek

> -----Original Message-----
> From: owner-po-dev at brie4.cshl.org 
> [mailto:owner-po-dev at brie4.cshl.org] On Behalf Of Toby Kellogg
> Sent: Monday, March 01, 2004 5:02 PM
> To: po-dev at plantontology.org
> Subject: Re: progress and thought questions
> 
> 
> Thanks Pankaj -
>   Just a couple of quick responses while I am thinking of it:
> >
> >> 7.  Is it worth including perigynium as an instance of a 
> prophyll?  
> >> The term is used only for species in the family Cyperaceae, 
> >> specifically species of Carex.
> >
> >My argument would be to include the terms on usage/requirement basis 
> >unless absolutely necessary to clarify some other node in 
> the ontology.
> 
> this sounds like a good principle to me.

XXX While I appreciate the probable pragmatism associated with this
principle I suggest that we need to be more proactive in these early
days of our 'product'. I suspect that our potential users may be more
sophisticated in their needs than we are perhaps currently anticipating.
I offer further comment below....
> 
> >
> >> 10.  Inflorescence can be defined with panicle, raceme, 
> tassel, ear, 
> >> and cob as synonyms.  Alternatively, we can divide inflorescences 
> >> into racemes (which do not terminate in a flower) and cymes (which 
> >> do).  Panicle, tassel, ear, and cob would thus be synonyms 
> of raceme.  
> >> Or we can keep dividing the inflorescence categories more and more 
> >> finely depending on which axes end in flowers and which don't, and 
> >> could end up with a plethora of terms.  (Same issue as #s 
> 2, 4, and 
> >> 8.)  How complex do we get?
> >
> >I agree with you. My question here is how do we deal with terms like 
> >primary/secondary inflorescence branches? There are many 
> mutants/genes 
> >responsible for the branching organization.
> 
> This gets back to the goal of the ontology.  For the many 
> mutants/genes that affect inflorescence branching, do we need 
> to be able to describe each of them in detail (in which case 
> we need lots of terms) or just to assign them to the bin 
> "inflorescence" (in which case we only need a few terms)? I 
> don't know the answer to this, but it will affect the 
> structure I think. i.e., will we want to make queries such as 
> "find all genes in any plant that, when mutated, affect 
> phyllotaxis on third order inflorescence branches", or will 
> it be sufficient to ask "what genes are expressed in inflorescences"?
>   Food for thought anyway.
> Toby
> 

XXX To continue, the above example is seemingly very appropriate. Yes, I
think we do need to provide the depth of information on all nodes so
that detailed assigning can be done - so we do need to provide lots of
terms. Providing the level of detail which just enables folk to assign
info. to a very course bin (like 'inflorescence') is surely rather
inadequate, considering the level of the science i.e. that many users
are working at a much finer level of granularity. Furthermore, I think
it is only when we 'sweat' at developing the finer structure now, will
we encounter & overcome the 'structure difficulties' that are possibly
lurking. This 'building' is going to be complex/multi-floored... -we
need to ensure that we've built the foundation well so that the
super-structure can be supported. While retrofitting should always be
possible I urge us all to go deeper now rather than leaving that for
later. To me this detailed approach is/should be the goal of the
'consensus ontology'. Perhaps, as a compromise, we could develop certain
nodes with fine granularity, based on perceived needs of our potential
consumers - here I'm assuming that some areas of anatomy are receiving
much more research attention than other (perhaps a wrong assumption).
Other nodes, perceived to be 'less needful' of fine granularity could be
worked on 'later' - but we should advertise that users should voice
their need for increased granularity wherever such needs are
encountered.

- Leszek


> Elizabeth A. Kellogg
> Department of Biology
> University of Missouri-St. Louis
> 8001 Natural Bridge Road
> St. Louis, MO 63121
> phone: 314-516-6217
> fax: 314-516-6233 
> http://www.umsl.edu/divisions/artscience/biolo> gy/Kellogg/Kellogg/
> 
> 
>