Re: [Obo-format] Latest obo2owl mapping

[Alan R. <ala...@gm...>]

On Jun 30, 2006, at 8:29 PM, chris mungall wrote:

> 2. Protege now seems to be classifying the ontology as OWL Full
>
> Can anything be done about 2?

I tried to generate some OWL in the new format but I'm not sure that  
I am seeing the right thing. What I see has a number of issues that  
would make it OWL full for no good reason, e.g.

  - Class and property definitions outside the ontology body
  - use of rdf:property to define some properties
  - namespace issues - e.g. obo namespace defined as http:// 
www.geneontology.org/formats/obo#, but base is "http:// 
www.bioontologies.org/2006/02/obo. Properties are defined with  
rdf:ID="blah" putting them in base, but then referenced as <obo:blah,  
which is a different.
  - Thing that look like they are intended to be anonymous  
individuals (but not properly formed for OWL-DL if they are)
  - ranges on what look like annotation properties

Anyways, to be sure I am looking at the right thing, could you send  
me a snippet of the generated owl please? Everything from the top to  
about 10 GO term class definitions down should be enough. I'll see if  
I am seeing the same thing and if so massage it into my best guess as  
to what was intended.

Regards,
Alan