Dear Nina,

> >
> > Up to now, we have been using ot:isA in the metadata to characterize
> > models, features, algorithms etc and rdf:type to assign OpenTox classes.
> > Ambit seems to use only rdf:type (at least in Features) and I have the
> > impression that ot:isA is redundant to rdf:type. Is that true?
> >
> 
> 
> Yes, rdf:type has meaning of "is a a type of <certain class> indeed and is a
> standard construct, contrary to OT- specific property  isA.
> 
> 
> >
> > If yes: Can we remove ot:isA from the API (I did not check if it is still
> > there) and deprecate ot:isA?
> >
> 
> Yes, I have indeed suggested this some time ago to the list and wiki, that
> ot:isA should be replaced by the standard RDF construct of subclassing.  So
> far I think TUM had also started using rdf:type instead of ot:isA.
> 
> Let's have ot:isA officially deprecated.

I have replaced ot:isA in the API 1.2 documentation, but not in the RDF
documentation (http://opentox.org/search?SearchableText=isA) and in
opentox.owl - can you fix it there?

Future versions of our services will use (and rely on) rdf:type instead
of ot:isA.

Best regards,
Christoph