=Paper= {{Paper |id=None |storemode=property |title=Modular Extensions to the ChEBI Ontology |pdfUrl=https://ceur-ws.org/Vol-897/poster_7.pdf |volume=Vol-897 |dblpUrl=https://dblp.org/rec/conf/icbo/HastingsMDEMTOS12 }} ==Modular Extensions to the ChEBI Ontology== https://ceur-ws.org/Vol-897/poster_7.pdf
                         Modular Extensions to the ChEBI Ontology
         Janna Hastings∗, Paula de Matos, Adriano Dekker, Marcus Ennis, Venkatesh
            Muthukrishnan, Steve Turner, Gareth Owen and Christoph Steinbeck
                     Cheminformatics and Metabolism, European Bioinformatics Institute, Hinxton, UK




ABSTRACT                                                                  from the core database content. The mapping to BFO is made
   ChEBI is an ontology of chemical entities of biological interest. It   available as one such modular extension, chebi-in-bfo.owl.
includes a structure-based and a role-based classification of chemical       Another extension, available as chebi-disjoints.owl,
entities including molecules, ions and salts. This poster showcases       includes axioms about disjointness between entities in the onto-
recent developments and forthcoming enhancements to the ontology,         logy. Many of the classes in the chemical entity ontology are not
focusing on modular extensions of the content. These extensions           disjoint, because chemical classification is compositional (Hastings
are being developed in support of error detection and consistency         et al., 2012). Annotating those pairs of classes that are disjoint thus
checking and to improve the capability for automated reasoning about      allows for better error detection, as a reasoner is able to raise an error
the biological context of chemicals.                                      if anything is accidentally classified beneath both of the two clas-
                                                                          ses that should be disjoint. In an ontology the size of ChEBI, such
1   INTRODUCTION                                                          automated assistance with the detection of errors is essential. For
                                                                          example, group (CHEBI:24433) is disjoint from molecular entity
ChEBI is an ontology of chemical entities of biological interest
                                                                          (CHEBI:23367). Introducing this axiom allowed us to detect and
(de Matos et al., 2010). It includes a structure-based and a role-
                                                                          correct an incorrect classification of the ion 3-D-glucuronosyl-N2,6-
based classification of chemical entities, such as molecules, ions
                                                                          disulfonato-beta-D-glucosamine(3-) (CHEBI:58150) as a group.
and salts. Within the structure-based classification, chemical enti-
ties are organised according to shared structural features, such as
skeletons and functional groups. In addition to taxonomic classi-         4    THE BIOLOGICAL CONTEXT OF CHEMICALS
fication, structure-based relationships capture biologically relevant     ChEBI is also being extended in order to explicitly represent the
associations between compounds, such as tautomers, enantiomers            biological context of the chemicals as semantic relationships to bio-
and conjugate bases and acids. Within the role classification, the        logical entities in external ontologies, as described in (Batchelor
various ways that chemical entities can be active in biological and       et al., 2010). One example of such is the explicit annotation of spe-
chemical contexts are organised, such as by inhibiting the activity of    cies and tissue for metabolites (natural products). Another is the
enzymes, acting as hormones, being used as insecticides, and being        explicit annotation of diseases in the case of drugs used to treat the
used as solvents. Chemical entities are linked to roles with the has      diseases and in the case of metabolites that are markers for certain
role relationship.                                                        diseases. Cross-references and annotations of ChEBI identifiers in
   ChEBI includes just over 28,000 annotated entities as of the           biological databases will be used as sources for this effort, which
May 2012 release. The ontology is available for browsing at               will then be checked by ChEBI curators.
http://www.ebi.ac.uk/chebi/ and is fully downloadable                        Relationships from the RO and other ontologies from the OBO
in several formats including OBO and OWL.                                 Foundry effort will be used in all cases.

2   RECENT DEVELOPMENTS                                                   ACKNOWLEDGEMENTS
ChEBI recently completed a large annotation push to make the onto-        ChEBI is funded by the BBSRC, grant agreement BB/G022747/1
logy fully is-a complete, that is, to ensure that each entity has at      within the ‘Bioinformatics and biological resources’ fund.
least one taxonomic parent. (Previously, due to historical legacy,
there were some entities that were only classified using structural       REFERENCES
relationships such as has functional parent.) Furthermore, ChEBI is       Batchelor, C., Hastings, J., and Steinbeck, C. (2010). Ontological dependence, dispo-
now aligned with BFO, as described in (Hastings et al., 2011) and            sitions and institutional reality in chemistry. In A. Galton and R. Mizoguchi,
work is well underway in alignment with the Gene Ontology (The               editors, Proceedings of the 6th Formal Ontology in Information Systems conference,
                                                                             Toronto, Canada.
Gene Ontology Consortium, 2000). Another recent development is
                                                                          de Matos, P., Alcántara, R., Dekker, A., Ennis, M., Hastings, J., Haug, K., Spiteri, I.,
the provision of a new graph-based visualisation of the ontology.            Turner, S., and Steinbeck, C. (2010). Chemical Entities of Biological Interest: an
                                                                             update. Nucleic Acids Research, 38, D249–D254.
3   EXTENSIONS FOR ERROR DETECTION                                        Hastings, J., de Matos, P., Dekker, A., Ennis, M., Haug, K., Josephs, Z., Owen, G.,
                                                                             Turner, S., and Steinbeck, C. (2011). Recent developments in the ChEBI ontology.
ChEBI is being enhanced with modular extensions to the onto-                 In Proceedings of the International Conference on Biomedical Ontology (ICBO),
logy in support of error detection and consistency checking. These           CEUR-WS volume 833.
extensions are being annotated in separate OWL module files which         Hastings, J., Magka, D., Batchelor, C., Duan, L., Stevens, R., Ennis, M., and Steinbeck,
                                                                             C. (2012). Structure-based classification and ontology in chemistry. Journal of
extend the main ontology file that is generated on a monthly basis
                                                                             Cheminformatics, 4(1), 8.
                                                                          The Gene Ontology Consortium (2000). Gene ontology: tool for the unification of
∗ To whom correspondence should be addressed: hastings@ebi.ac.uk             biology. Nat. Genet., 25, 25–9.




                                                                                                                                                                 1