Difference between revisions of "XP:Main Page"
Line 13: | Line 13: | ||
etc | etc | ||
− | The term "cross-product" comes from Hill et al's work on GO | + | The term "cross-product" comes from Hill et al's work on GO, which proposed rational combination of basic process terms with terms from anatomical ontologies. The Manchester GONG Project (http://gong.man.ac.uk/) proposed the use of OWL to create logical class expressions for GO terms, and the [http://www.berkeleybop.org/obol Obol] project set about trying to reverse engineer these logical definitions for existing OBO ontologies. Both these efforts placed importance on the relations used in the logical definition/cross product, spurring the development of the [http://www.obofoundry.org/ro OBO Relation Ontology]. |
There are various ongoing efforts to maintain these logical definitions. In some circumstance it is beneficial to employ "post-composition", in which case the composition is made at annotation time, rather than prior to ontology deployment. From a logical point of view, these are equivalent. | There are various ongoing efforts to maintain these logical definitions. In some circumstance it is beneficial to employ "post-composition", in which case the composition is made at annotation time, rather than prior to ontology deployment. From a logical point of view, these are equivalent. |
Revision as of 12:49, 20 July 2007
This subwiki is for the discussion and coordination of 'cross-products' between OBO Ontologies, primarily those in the OBO Foundry.
Introduction
Many OBO Foundry ontologies have terms with implicit references to terms from other ontologies:
- Diseases involve biological processes (GO) and dispositions, and are often manifested or arise from phenotypes (eg MP)
- Phenotypes are aggregates of qualities (PATO) in bearer entities, such as those represented in anatomical ontologies and GO
- Biological processes (GO) have continuants as participants, and these continuants are represented in ontologies such as ChEBI, GO-CC, Cell, anatomical ontologies, ...
- Terms in the GO-CC ontology can be refined according to the type of cell (CL) their instances are located in, and cell terms can be refined using terms from gross anatomical ontologies.
- Environments can be defined in terms of ranges of qualities found in certain zones
etc
The term "cross-product" comes from Hill et al's work on GO, which proposed rational combination of basic process terms with terms from anatomical ontologies. The Manchester GONG Project (http://gong.man.ac.uk/) proposed the use of OWL to create logical class expressions for GO terms, and the Obol project set about trying to reverse engineer these logical definitions for existing OBO ontologies. Both these efforts placed importance on the relations used in the logical definition/cross product, spurring the development of the OBO Relation Ontology.
There are various ongoing efforts to maintain these logical definitions. In some circumstance it is beneficial to employ "post-composition", in which case the composition is made at annotation time, rather than prior to ontology deployment. From a logical point of view, these are equivalent.
Some outstanding issues include species-specificity, defining relations correctly, and handling inter-ontology dependence
Resources
The majority of the work done so far has been in the GO. See the BBOP wiki page for coordinating efforts on logical definitions for GO terms that refer to cell types:
this page also has a summary of other potnential OBO Foundry cross-products
The sequence ontology already has intra-ontology cross products (also done as part of GO Consortium efforts):
For cross-products involving phenotype ontologies, see:
Downloading and using
For a list of vetted cross-products plus availability in OBO and OWL format, see
For a wider list of automatically generated cross-products, see the Obol pages: