Re: [Obo-phenotype] quick question: PATO & mereotopological relations

[William B. <William.Bug@DrexelMed.edu>]

On Apr 20, 2007, at 3:47 PM, Melissa Haendel wrote:

> ZFIN is currently using the PATO spatial relations to indicate that  
> something is changed relative to normal, not to indicate normal  
> position.  I think we'd want to determine if these two uses would  
> require relations with the same definition or not.

This is an excellent point, Melissa, and one I'm not at all certain  
how to call.  It's likely in the long run, it will be something like  
the following:
	- spatio-temporal detail on the "normal" state will be represented  
using a mereotologically complete and consistent set of relations  
(perhaps from OBO RO - or an anatomically-focussed OBO RO derivative)
	- the "non-normal" state will then be described using PATO IN  
REFERENCE TO the complete description of the normal state.

To some extent, this over-simplifies the complexity of the  
representational task, but I do think something close to this will  
ultimately be possible.

In the short term, the solution you currently have sounds like a very  
practical one.


On Apr 20, 2007, at 3:47 PM, Melissa Haendel wrote:
> The FMA has a set of spatial relations as non-physical anatomical  
> entities. When building the first release of CARO, we decided that  
> having the spatial relations in an anatomy ontology of anatomical  
> entities was somewhat inconsistent and so we left them out despite  
> using the FMA as a model for CARO.

I agree with this opinion whole-heartedly.  In the BIRN ontology  
(BIRNLex), we have derived a representation of the very comprehensive  
NeuroNames neuroanatomical concept map as a subsumptive ("is_a")  
graph of volumetric entities (e.g., "abducens nucleus" is_a  
"regional_part_of_pons").  We are using OBO RO derived relations to  
then specify the mereotopological detail.  To date, it appears to be  
working well, and provides us a means to both capture those aspect of  
detailed brain anatomy that all agree on, and a way to incorporate  
distinct views that diverge at the more granular spatial level -  
though this is still very much a work-in-progress.

In the end, if the mereotopological relations we use in BIRNLex ONLY  
have meaning within BIRNLex, we won't have really fulfilled our  
mandate to produce something of use to the broader community.

Our next goal is to make certain we can correctly map to CARO, which  
SHOULD be made a little easier by the fact we've built everything off  
of the Basic Formal Ontology.

Cheers,
Bill


Bill Bug
Senior Research Analyst/Ontological Engineer

Laboratory for Bioimaging  & Anatomical Informatics
www.neuroterrain.org
Department of Neurobiology & Anatomy
Drexel University College of Medicine
2900 Queen Lane
Philadelphia, PA    19129
215 991 8430 (ph)
610 457 0443 (mobile)
215 843 9367 (fax)


Please Note: I now have a new email - William.Bug@DrexelMed.edu