search

nvaldivi / ogms

Automatically exported from code.google.com/p/ogms
0 stars 0 forks source link

issues and questions from Chris Mungall #39

Open GoogleCodeExporter opened 8 years ago

GoogleCodeExporter commented 8 years ago 
"predisposition to disease of type X"
is an odd name for a class. Either this is a template - in which case it 
shouldn't be represented as 
a class in the ontology - or the class is actually "predisposition to disease"

The paper mentions relations sign_of and symptom_of - should these be in here?

syndromes are ICEs?

Should there be an axiom linking abnormal homeostasis to clinically abnormal?

shouldn't disease phenotype be a subtype of clinical phenotype (based on the 
definition).

Are there disease phenotypes that are not clinical phenotypes? What about plant 
diseases?

[these should be split out into separate issues]

Original issue reported on code.google.com by alanruttenberg@gmail.com on 7 Dec 2009 at 5:06

GoogleCodeExporter commented 8 years ago 
Chris:

  "predisposition to disease of type X"
  is an odd name for a class. Either this is a template - in which case it
  shouldn't be represented as a class in the ontology - or the class is
  actually "predisposition to disease"

Barry:

 The rationale here is that IDO is a core ontology, which will exist in its
 full form only in the various extension ontologies -- IDO-Malaria, IDO-HIV,
 etc. -- where the 'X' will be replaced by the name of some specific disease

Alan:

Chris' point still stands. This isn't a real class - it's a template
for creating a class. We need to figure out a way of clearly
distinguishing such things from bona fide universals.

This sort of thing is one of the reason that FMA has templates and
uses protege frames.

I think we need a standard way of referring to a class definition as
an information artifact rather than as the "see-through" universal. A
template restricts the form of the information artifact (which of
course affects the denotation as well).

Either that, or perhaps have the class be a defined class:
"predisposition to a disease" and have an editor note to the effect
that the desired differentia for immediate subclasses is the specific
disease. Of course, since it would be a defined class, one wouldn't
*need* to assert the subclasses as subclasses.

Original comment by alanruttenberg@gmail.com on 7 Dec 2009 at 5:07