Feedback on POC (fwd) (PR#89)

[François Delcoux]

Hello!

I wrote to you a couple of weeks ago about some bugs I found in PO.

Well, now, I propose making a kind of a "deal" with you...

See, I am an undergraduate student in Plant Molecular Biology and my  
end-of-study work is all about making bio-ontologies user-friendly.  
My work takes part in a universal, versatile bioinformatics framework  
that uses OWL ontologies (since they're more universal then OBO ones).
So I had to write a toolbox in order to get OWL files directly from  
OBO flat files (instead of using the OBO-xml intermediate with  
currently available XSLT scripts that just give awful results ! ).  
This toolbox is able to track some errors like two different entries  
having the same ID, for example, and is intended to be deployed on a  
web server in my university.

Here is the last report I got from my OBOmerge utility (since I need  
both PO_anatomy and PO_temporal at the same time and I need that  
merge to be as fresh as possible) :

> Trying to merge 'po_anatomy.obo' and 'po_temporal.obo'.
>
> I found different entries with the same ID (PO:0004500).
> I found different entries with the same name ("radicle emergence").
>
> Note: version of 'po_anatomy.obo' is not specified in the OBO file.

Would you mind if I set a routine in my toolbox that warns you  
automatically when an error is detected (once a week at the most)?  
Could you, please, give me a bug-reporting e-mail address for such a  
purpose?

I think that the OWL version of PlantOntology that we will maintain  
could help you dealing with missing 'is_a' relationships, since OWL  
has a very pure hierarchy (only 'subclass_of'/'is_a' relationships ;  
other relationships are considered as properties).

So there's the deal... Do you agree?

Besides, I noticed that you had Terms whose names matched those of  
the namespaces ('plant structure' and 'plant growth and development  
stage'). Now, 'plant growth and development stage' has been renamed  
to 'plant growth and development stages' as the name of the PO: 
0009012 Term, but the namespace is still in the singular form. Since  
I used those Terms as namespaces equivalents, It causes a bug in one  
of my progs... Please, tell me if you're going to maintain that  
little 'inconsistency' (so I will modify my prog) or to correct it.

In the hope I may be useful...
Yours sincerely,
François DELCOUX


Le 28 oct. 2005 à 02:56, Katica Ilic a écrit :

> Dear François,
>
> Thank you for bringing to our attention an existence of terms in  
> the PO with
> missing relationships. We recognized this problem some time ago and  
> have been
> trying to solve it. In fact, one of our colleagues, Pankaj Jaiswal,  
> has been
> actively working on adding 'is_a' relationships to PO terms. This  
> month, the
> ontology release will include updates for some of them, for  
> instance the term
> you mentioned 'adventitious root epidermis'. Since there are many  
> terms with
> missing 'is_a' relationship, it may take several rounds of ontology  
> updates to
> complete this task and have the ontology where all terms would have  
> 'is_a'
> relationship.
>
> Regarding your suggestion to include
> 'obsolete_plant_growth_and_development_stage', this is exactly what  
> we currently
> have in PO as the top grouping term for all obsoleted terms in this  
> aspect of
> PO. The term you were referring to is just and old term that was  
> obsoleted early
> on while we were creating PO. We don’t use, maintain or change  
> names of
> obsoleted terms, since they are in the ontology mainly for the  
> historical
> prospective and for keeping a record of all the terms entered in  
> the ontology
> (no term is ever deleted in PO).
>
> Thank you for using Plant Ontologies and for contacting us.
>
> Best regards,
>
> Katica Ilic, POC Project Coordinator
>
>> ---------- Forwarded message ----------
>> Date: Thu, 27 Oct 2005 10:45:39 -0400
>> From: feedback_submission at filetta.cshl.edu
>> Reply-To: po-dev at plantontology.org,  
>> feedback_submission at filetta.cshl.edu
>> To: po-dev <po-dev at plantontology.org>
>> Subject: Feedback on POC
>>
>>      *** Feedback from Plant Ontology Live Site ***
>>
>> refer_to_url: http://www.plantontology.org/index.html
>>
>> comments: Hello!
>>
>> I noticed that some obvious "is_a" relationships are missing...
>>
>> eg: Term "adventitious root epidermis" should have an "is_a"  
>> relationship to
>> "root epidermis";
>> Term "aleurone layer" should, at least (maybe not directly, but  
>> through some
>> sub-class) have an "is_a" relationship to "Tissue"; (as far as I  
>> know)
>>
>> Besides, I think that it would be better to rename the Term  
>> "obsolete growth
> and
>> development terms" to "obsolete plant growth and development  
>> stage", in order
> to
>> be consistent with its structural counterpart.
>>
>> Thanks for reading me.
>> Bye
>>
>> name: Fran篩s DELCOUX
>>
>> email: fdelcoux at ulb.ac.be
>>
>> organization: ULB (Belgium)
>>
>> send_feedback: Send your feedback
>> --1118555204-1981955417-1130439040=:13943--
>>
>>
>