Re: blog: semantic dissonance in uniprot

On Mar 25, 2009, at 12:30 PM, Oliver Ruebenacker wrote:

>     Hello Mark, All,
>
> On Wed, Mar 25, 2009 at 1:13 PM, Mark Wilkinson  
> <markw@illuminae.com> wrote:
>> Well, the statement would *imply* that it is... so given that the  
>> individual
>> "embryo" that was referred to as a uniprot tissue is the same  
>> individual
>> "embryo" that the plant ontology was talking about, we can therefore
>> conclude that (as Ben pointed-out) that this particular fly embryo is
>> somehow embedded in some particular plant seed endosperm.
>
>  Just because it refers to a set of things does not mean I need to
> model it by an owl:Class.

No, but if it really is a set, that would be a very good idea.

> I can also have an owl:Class Population and
> an instance of it populationOfTheUS, which would be an instance, not
> an owl:Class, although it seems to correspond to a set of people.

Well, it was a problem with OWL DL that an class couldnt be an  
instance of another class. I gather this has been fixed (about time)  
in OWL 2,  There is obsolutely no logical or fundamental reason why  
classes should not be allowed to be in other classes, for examples  
like this one.

> Similarly, a typical approach would be to have a class Protein and an
> instance EGFR ("a protein") that refers to a large number of molecules
> scattered all over the globe.
>

Um.. this seems confused to me, mixing up ideas from mereology with  
classes. A 'large number of molecules scattered all over the globe'  
sounds like a mereological sum. Which is fine, and pretty well  
axiomatized already with off-the-shelf ontologies. BUt there is no  
point in also having a class of these things, since this just confuses  
sums with classes. What are the elements of the class? The sum or the  
individual molecules? If the latter, we don't need the sum; if the  
former, we don't need the class.

>  The statements from UniProt, as I understand them, make perfect
> sense: there is a protein that has been isolated from that plant, and
> can also be found in fruit fly embryos. I don't have the knowledge to
> check whether it is true, but it seems to make sense to me.

I agree it makes sense. But to be exact, whats been found in fruit fly  
embyos are molecules of the (same) protein, not the same molecules of  
the protein. If this is important for reasoning (and I bet it will  
be), then your ontology needs to distinguish protein molecules from  
proteins. And if it doesn't, then I'd predict it is going to get its  
knickers in a twist.

Pat

>
>     Take care
>     Oliver
>
> -- 
> Oliver Ruebenacker, Computational Cell Biologist
> BioPAX Integration at Virtual Cell (http://vcell.org/biopax)
> Center for Cell Analysis and Modeling
> http://www.oliver.curiousworld.org
>
>
>

------------------------------------------------------------
IHMC                                     (850)434 8903 or (650)494 3973
40 South Alcaniz St.           (850)202 4416   office
Pensacola                            (850)202 4440   fax
FL 32502                              (850)291 0667   mobile
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes

Received on Thursday, 26 March 2009 07:30:12 UTC