BioPAX-OBO-Memorandum

This is the working draft -

On Nov 20-22 members of the BioPAX and OBO Foundry communities[1] participated in a workshop focused on ontological issues in pathway representation. This memorandum is intended to record some consensus understandings that were reached. No commitments were made by either BioPAX or OBO-F.

Similarities between OBO-F and BioPAX efforts:


 * Both communities share a commitment to the development of standardized, machine understandable models of biology.
 * Both communities are interested in going beyond terminological hierarchies to explore richer structured representations based on formal knowledge representation languages such as description logics, particularly the semantic web’s OWL-DL language.
 * Both communities face common technical problems relating to ontology integration, scalability and the relation between ontologies and databases.

Potential interactions between BioPAX and OBO-F:
 * What can OBO-F do for BioPAX? OBO-F component ontologies are intended to standardize terms that are in the BioPAX domain of interest, including GO-CC, GO-BP, and others. The use of such ontologies is consistent with BioPAX’s commitment to reuse existing standards, and supports ontology re-use envisioned for the semantic web.
 * What can BioPAX do for OBO-F? Members of the BioPAX community have experience with the issue of mapping pathway knowledge to description logics and OWL, and could serve OBO-F as an “advance guard” in this effort. We believe that it would be fruitful to explore the possibility of using the BioPAX initiative as a forum for using and integrating ontology resources and for testing the utility of ontologies for supporting automated reasoning in the life sciences.
 * What can BioPAX and OBO do for each other?
 * Share use cases and solutions to illustrate ontology design patterns for modeling complex biological phenomena: molecular processes and states (protein phosphorylation, etc.), processes and events, causality, the issue of provenance of data and the relationship between pathway representations and experimental and phenotypical observations.
 * BioPAX could be the responsible party that creates the formal definitions for GO biological process terms in the domain of molecular biology. The GO Consortium would then have to find a way to give credit to the BioPAX community and curators of pathway knowledge. [ One way to achieve this would be to create an editorial board for the GO-Pathway sub-ontology. ]

Barry Smith slides: http://ontology.buffalo.edu/bio/BioPAX_06

Cc: Biopax, go-friends, semweb lists

Dan Corwin, Emek Demir, Michel Dumontier, Ray Hookway, Cartik R. Kothari, Stan Letovsky, Joanne Luciano, Eric Neumann, Elgar Pichler, Jonathan Rees, Alan Ruttenberg, Barry Smith, Andrea Splendiani, Jeremy Zucker.

[1] The Semantic Web and NCBO were represented as well