10.14219/jada.archive.2010.00312010;141(10):1173-1175JADA

and Michael GlickBarry Smith, Louis J. Goldberg, Alan RuttenbergInformaticsOntology and the Future of Dental Research

jada.ada.org (this information is current as of August 17, 2014):The following resources related to this article are available online at

http://jada.ada.org/content/141/10/1173in the online version of this article at:

including high-resolution figures, can be foundUpdated information and services

http://jada.ada.org/content/141/10/1173/#BIBL, 0 of which can be accessed free:5 articlesThis article cites

http://www.ada.org/990.aspxthis article in whole or in part can be found at: of this article or about permission to reproducereprintsInformation about obtaining

prohibited without prior written permission of the American Dental Association. Copyright © 2014 American Dental Association. All rights reserved. Reproduction or republication strictly

on August 17, 2014jada.ada.orgDownloaded from on August 17, 2014jada.ada.orgDownloaded from

How do we find what is clinically significant in theswarms of data being generated by today’s diagnostictechnologies? As electronic records become ever moreprevalent—and digital imaging and genomic, proteomic,salivaomics, metabalomics, pharmacogenomics, phe-

nomics and transcriptomics techniques become commonplace—different clinical and biological disciplines are facing up to the needto put their data houses in order to avoid the consequences of anuncontrolled explosion of different ways of describing information.

Fortunately, a new strategy to advance the consistency of data inthe dental research community is emerging. The strategy is basedon the idea that existing systems for data collection in dental re-search will continue to be used, but proposes a methodology inwhich past, present and future data will be described using a consensus-based controlled structured vocabulary called theOntology for Dental Research (ODR). The ODR initiative is modeledon a series of existing biomedical ontology projects and will adoptbest-practice principles that already have been thoroughly tested inareas such as molecular biology, model organism research, pro-teomics and genetic disease.1

An “ontology,” in this context, is a controlled, logically structuredvocabulary created by experts in a given area as a strategy for pro-moting consistency in the way primary data (for example, in theform of experimental results or clinical records) are described.Specialist biocurators create “annotations” in the form of HTML tagslinking such primary data to expressions in the ontology, therebymaking the data available to search and to algorithmic processing.

Each ontology contains a taxonomy at its heart, and its logicalstructure is built around the hierarchy defined by its taxonomic(subtype) relation. But an ontology contains also definitions of itsterms, along with additional relations such as parthood, connectionand participation, as well as functional relations. These additionalrelations make the data searchable not only through the use ofterms in the ontology, but also through logically related terms.Thus, the ontology can be used to retrieve data associated with

Over time, more andmore researchers will

see the value of employing a common

resource in annotatingtheir data and in

designing new databases in which

to capture their research results.

Barry Smith, PhD; Louis J. Goldberg, DDS, PhD;Alan Ruttenberg; Michael Glick, DMD

C O M M E N T A R Y E D I T O R I A L

JADA 141(10) http://jada.ada.org October 2010 1173

ED ITORIAL

Ontology and the future ofdental research informatics

Editorials represent the opinions of the authors andnot necessarily those of the American Dental Association.

Copyright © 2010 American Dental Association. All rights reserved. Reprinted by permission.

on August 17, 2014jada.ada.orgDownloaded from

terms referring to parts of spe-cific anatomical entities, toanatomical entities immediatelyconnected to specific anatomicalentities, or to biological process-es in which specific anatomicalentities participate.

We can conceive ODR, in thefirst place, as providing anevolving standard set of keywords for all aspects of dentalresearch. Initially, these keywords can be used to annotateboth published literature andexisting research databases.Such annotation will enableeasier access to research resultsand allow also first steps towardthe semantically enhanced pub-lishing of the future.2 The long-term goals of ODR, however,are much more ambitious. ODRwill include not only English-language definitions of its termsfor human use and for humanquality control of the ontology,but also logical definitions foruse by computers. With the lat-ter, ODR then can be used as acomputational resource for en-hanced search and integrationof data, and for reasoning—notonly with dental research databut also with data annotatedusing other biological and bio-medical ontologies with whichODR will be linked logically.

The key idea behind ODR isrooted in 10 years of experienceusing ontologies in support ofbiomedical research. Ontologiesin biomedicine began in themodel organism community,which faced a problem of incon-sistency in the ways in whichthe results of functional ge-nomics experiments on differentkinds of organisms were beingdescribed. To address theseproblems, a group of leadingmodel organism databases cametogether in 1999 to create theGene Ontology (GO), a con-

trolled structured vocabularyfor describing different attribut-es of gene products.3

The GO is designed to bespecies neutral. It provides a setof some 30,000 common termsfor describing different kinds ofcellular constituents, biologicalprocesses and molecular func-tions in all kinds of organisms—terms such as “mitochondrion”or “cell division” or “binding.”Since its inception, more than$100 million has been investedin the use of the GO to annotatereferences to gene products indatabases and in the scientificliterature. There are more than11 million annotations relatinggene products described in theUniProt, Ensembl and otherdatabases and in more than50,000 scientific journal articlesto terms in the GO.4 The infor-mation in huge numbers of dis-persed resources is hereby beingmade accessible through re-sources such as AmiGO andGOPubMed. Increasingly, theavailability of this huge body ofintegrated information also ishaving an influence on clinicalresearch, and a simple PubMedsearch on “gene ontology” re-veals a variety of different waysin which the GO and the dataannotated in its terms are beingused in support of research onhuman health and disease.

Important features of ODRinclude the following:dIt will be built to workwith the GO and with otherhigh-quality ontologies de-veloped by the biomedicalcommunity. This means thatODR will follow the best prac-tices identified through 10 yearsof testing by the GO and by itssister ontologies participating inthe Open Biomedical Ontologies(OBO) Foundry initiative.5

dIt will be built with terms

used by dental researchers,and it will be created andmanaged by the dental re-search community itself. Themore an ontology is used, themore the ontology and the datadescribed in its terms increasein value, and the more researchgroups in the future will be mo-tivated to use the ontology indescribing their data. The keyto ontology success, therefore, isincentivizing users, and to thisend it is important that poten-tial users feel that they haveownership of the ontology, thatthe ontology is populated usingthe terms that they need anduses definitions that conform totheir understanding of theseterms. ODR is being initiated bythe leading informaticiangroups within the dental re-search community in such away that it will, from the verystart, be in a position to serveas an attractor for multiple ex-panding groups of users whosemembers will have strong in-centives not only to invest re-sources directed toward ensur-ing that it is developed in waysthat keep pace with scientificadvance, but also to recommendit to other users—thereby in-creasing the value of their owninvestment in the resource.dIt can be corrected easilyin light of new research dis-coveries. One key presupposi-tion for the success of an ontol-ogy project is its ability tointegrate previously annotateddata with new terms and rela-tions brought to light by ongo-ing scientific discovery. Thisprocess ensures that previouslyannotated (legacy) data do notlose their value. To this end, thebiomedical ontology communityhas developed a methodologybased on careful versioning ofontologies and annotations, com-

C O M M E N T A R Y E D I T O R I A L

1174 JADA 141(10) http://jada.ada.org October 2010

Copyright © 2010 American Dental Association. All rights reserved. Reprinted by permission.

on August 17, 2014jada.ada.orgDownloaded from

bined with software tools to en-sure consistent updating of exist-ing annotation resources witheach new version of the ontology.dIt can be extended easilyto incorporate new kinds ofdata. The organization of OBOontologies is based on the use ofa simple and highly flexible tree-like hierarchy structure. Thiscan be extended at will to com-prehend new domains of entitiesas science evolves, and therebyallow the annotation of newkinds of data in ways consistentwith existing annotations.

The ODR will benefit the re-search community in a numberof ways: dIt is designed to work wellwith existing ontologies in allareas of clinical and translation-al science, and thus allows den-tal research data to be easily in-tegrated with other kinds ofdata.dIt is designed to work wellwith the Semantic Web, provid-ing access to all data resourcesthrough unique Web URLs asso-ciated with each ontology term.6

dIt provides a pretested andwell-defined set of terms, selec-tions from which can be used inthe design of new databases.dIt can incorporate, whereneeded, sets of synonyms deriv-ing from legacy term sets andnomenclatures such as theSystematized Nomenclature ofMedicine–Clinical Terms andSystematized Nomenclature ofDentistry vocabularies.7,8

dTo ensure high quality andcontinued maintenance, theODR controlled vocabulary willbe subject to a process of gover-nance and peer review.dOrganizations such as theNational Institutes of Healthare requiring definitions of com-

mon standards to ensure thatthe results obtained throughfunded research are more easilyaccessible to external groups.ODR will be created in such away that its use will meet thesecommon standards. It is de-signed also to allow informationpresented in its terms to be us-able in satisfying regulatorypurposes—submissions to theU.S. Food and Drug Adminis -tration, for example.

ORD will contain severalsubontology components, in-cluding the Salivaomics Ontol -ogy, 9 a Dental Anatomy Ontol -ogy based on the FoundationalModel of Anatomy10 and an OralPathology Ontology. In addi-tion, vocabulary resources arebeing developed, based on theOntology for General MedicalScience (OGMS)11 and theOntology for BiomedicalInvestigations,12 to representdental disease and dental pro-cedures, and to allow a seam-less connection between the useof ODR in the dental domainand the use of existing ontologyresources developed in otherareas of medicine.

The use of ODR to describedata will be entirely voluntary.However, we anticipate thatover time, more and more re-searchers will see the value ofemploying a common resourceboth in annotating their dataand, progressively, in designingnew databases in which to cap-ture their research results. ■

Dr. Smith is a State University of New YorkDistinguished Professor and the director,National Center for Ontological Research,Department of Philosophy, University atBuffalo, N.Y.

Dr. Goldberg is a professor, Oral DiagnosticSciences, School of Dental Medicine,University at Buffalo, N.Y.

Mr. Ruttenberg is a principal scientist,

Science Commons, Cambridge, Mass.

Dr. Glick is the dean, School of DentalMedicine, University at Buffalo, N.Y. He alsois the editor of The Journal of the AmericanDental Association. Address reprint requeststo Dr. Glick at School of Dental Medicine,University at Buffalo, 325 Squire Hall,Buffalo, N.Y. 14214-8006, e-mail“glickm@ada.org”.

1. Smith B, Ashburner M, Rosse C, et al.The OBO Foundry: Coordinated evolution ofontologies to support biomedical data integra-tion. Nat Biotechnol 2007;25(11):1251-1255.

2. Kidd R. Semantic enrichment boosts in-formation retrieval. Res Inform 2007; April/May. “www.researchinformation.info/features/feature.php?feature_id=127”.Accessed Aug. 25, 2010.

3. Ashburner M, Ball CA, Blake JA, et al.Gene ontology: tool for the unification of biology—The Gene Ontology Consortium. Nat Genetics 2000;25(1):25-29.

4. European Molecular Biology Laboratory,European Bioinformatics Institute. GeneOntology Annotation (UniProtKB-GOA)Database. “www.ebi.ac.uk/GOA/”. AccessedAug. 25, 2010.

5. The Open Biological and BiomedicalOntologies. “www.obofoundry.org”. AccessedAug. 25, 2010.

6. Ruttenberg A, Clark T, Bug W, et al.Advancing translational research with the se-mantic Web. BMC Bioinform 2007;8(suppl3):S2. “http://www.biomedcentral.com/1471-2105/8/S3/S2”. Accessed Aug. 25, 2010.

7. International Health TerminologyStandards Development Organisation.Welcome to IHTSDO. “www.ihtsdo.org/”.Accessed Aug. 31, 2010.

8. Goldberg LJ, Ceusters W, Eisner J, SmithB. The significance of SNODENT. StudHealth Technol Inform 2005;116:737-742.

9. Ai J, Smith B, Wong D. Saliva ontology:an ontology-based framework for a sali-vaomics knowledge base. BMC Bioinform2010;11:302. “www.biomedcentral.com/1471-2105/11/302”. Accessed Aug. 28, 2010.

10. Rosse C, Mejino JLV. The foundationalmodel of anatomy ontology. In: Burger A,Davidson D, Baldock R, eds. AnatomyOntologies for Bioinformatics: Principles andPractice. London: Springer; 2007:59-117.

11. Ontology for General Medical Science(OGMS). “www.acsu.buffalo.edu/~ag33/ogms.html”. Accessed Aug. 28, 2010.

12. The Ontology for BiomedicalInvestigations. “http://obi-ontology.org/page/Main_Page”. Accessed Aug. 25, 2010.

Additional ReadingsBiomedical ontology. “http://ontology.

buffalo.edu/biomedical.htm”. Accessed Aug.31, 2010.

Stevens R. Ontologies in biology. “www.cs.man.ac.uk/~stevensr/menupages/background.php”. Accessed Aug. 31, 2010.

Introduction to biomedical ontologies.“http://bioontology.org/wiki/index.php/Introduction_to_Biomedical_Ontologies”.Accessed Aug. 31. 2010.

Cambridge Healthtech Institute.Pharmaceutical ontologies & taxonomies glos-sary & taxonomy. “www.genomicglossaries.com/content/ontologies.asp”. Accessed Aug.31, 2010.

C O M M E N T A R Y E D I T O R I A L

JADA 141(10) http://jada.ada.org October 2010 1175

Copyright © 2010 American Dental Association. All rights reserved. Reprinted by permission.

on August 17, 2014jada.ada.orgDownloaded from