Human Ontology Resources

Our group have started to put together an ontology describing anatomical and physiological relationships. The reason for this is to integrate our CellML model repositories into an overall knowledge base which:

1. enables people to easily locate models through ontology relationships.
2. enables us to extend CellML into a more powerful conceptual modelling tool.
3. provides a framework in which to interpret, extend, and share physiological models.

At present the anatomical ontology work is very much in its infancy. We are using Protégé to experiment with different relationships and to help explore various tools for interacting with and processing ontology data.

The ontologies we are currently working on are anatomical relationships between blood vessels, bones, muscles and nerves; cellular processes such as reaction pathways, electrophysiological models, and developmental processes, and structural relationships such as organ and cell composition.

The CBIL Controlled Vocabulary for Anatomy provides a set of hierarchical terms for mouse and human anatomical descriptions initially based on GXD mouse stage 28 terms that have been extended to incorporate human anatomy and expanded particularly in
the hematolymphoid and central nervous systems.

The CBIL Controlled Vocabulary is freely available at
http://www.cbil.upenn.edu/downloads/RAD/ in 5 files
(controlled-vocab-*.sql) to create and populate tables for a relational
database.

eVOC is a set of four orthogonal controlled vocabularies which aim to provide an appropriately detailed set of terms for describing the sample source of human cDNA and SAGE libraries, and labelled target cDNAs for microarray experiments - and thereby for describing the location and timing of gene expression. The expression ontologies are independent pure hierarchies (or trees). In these pure hierarchies there is only a single type of relationship between the nodes in each hierarchy, although the nature of the relationship is not explicitly defined. For each ontology, the nature of the expression domain imposes an implicit type on the relationship between the nodes. For instance, in the "Anatomical System" ontology, the relationships are of the "part-of" type. In the "Cell Type" and "Pathology" ontologies, they are of the "subclass" type, and in the "Developmental Stage" ontology, the relationships are of the "is-a" variety. Definitions of each of the terms are to be provided as part of the ongoing development. The source of each definition will be made available along with the definition.

eVOC is provided under a BSD-style license and is available for download free of charge from http://www.sanbi.ac.za/evoc/ and can be used and modified without restriction. The ontologies are currently provided as flat text files. Work towards providing the ontologies in XML is underway.

From the website users are also able to download the annotated datasets, join the Expression Vocabulary Consortium and sign-up to use the eVOC mailing list. We maintain a central, versioned database of eVOC ontologies which are updated, modified and released publicly, by domain-experts on an ongoing basis.

Groups that choose to modify the ontologies for their own purposes are encouraged to contribute their modifications and corrections to the curators for inclusion. A mailing list: evoc@sanbi.ac.za has been established for this purpose.

The Foundational Model of Anatomy (FMA) is an evolving computer-based knowledge source for bioinformatics; it is concerned with the representation of concepts and relationships necessary for the symbolic modeling of the structure of the human body in a form that is understandable to humans and is also navigable by machine-based systems. Specifically, the FMA uses a frame-based formalism to represent a coherent body of knowledge about human anatomy as an ontology. The FMA contains approximately 70,000 concepts and over 100,000 terms; over 1 million iterations of 120 types of relationships link the FMA's concepts into a coherent symbolic model. It includes anatomic concepts and relations from the macromolecular to the organismic level. Although it is primarily a model for human anatomy, we believe it is largely generalizable to other mammals, and perhaps to all other vertebrates.

For questions that are entirely focused on the anatomy in our ontology, Onard will act as a secondary point of contact:
Jose L.V. (Onard) Mejino mejino@u.washington.edu

To register, interested users should send me an email with their name and affiliation. Please let us know what your goals might be for viewing the FMA, and also please give us your academic background (e.g. computer science, biology, genetics, etc...)

We have put together an ontology of human developmental anatomy for Carnegie stages 1-20. This ontology is modelled on that for mouse developmental anatomy (http://genex.hgu.mrc.ac.uk/Databases/Anatomy/) and currently only includes parts-of relationships. The anatomy is available in two formats:

The basic version This contains all the major named structures together with notes describing the detailed development and associated references. The database holding this information can be searched to give tissue lists together with first and last mentions of tissues.

The more complete version This contains the named tissues together with their component parts, and this version is thus at a finer resolution than the basic one. These lists area available at the above website, while a tree-based Java viewer is available at http://genex.hgu.mrc.ac.uk/Resources/intro.html.

The complete anatomy should also be available soon on the GOBO site http://www.geneontology.org/doc/gobo.html

NCI Thesaurus is a description logic cancer vocabulary developed using the Apelon Terminology Development Environment (TDE) and made available to applications through the NCI Terminology Server. Access to the NCI Thesaurus and other terminologies via Web browser is provided at http://nciterms.nci.nih.gov. See http://ncicb.nci.nih.gov/core/EVS for more detailed information.

Human Anatomy:

"Anatomic Structure, System or Substance" is structured as an Aristotelian hierarchy, organized by the is_a relationships, with physical_ part_of and has_location relationships (roles) to other concepts. If needed, a mixed hierarchy may be generated by combining the is_a and part_of and has_location relationships. In addition to macroanatomy concepts organized by Organ, System, Cavity, etc., the terminology includes microanatomy, embryologic structures and fluids at a granularity useful for biology research, and contains approximately 5500 concepts.

The human "Anatomic Structures, Systems" vocabulary is used for general purposes such as indexing documents, but also is used by the Cancer Models database, and the caIMAGE database in NCI. It has been designed to work well with the NCI Thesaurus disease terminology. It was reviewed by the Air Force Institute of Pathology and has been adopted as a CHI standard for research use by the United States federal government.

Activities in Process:

Through the Cancer Bioinformatics Grid project (caBIG) we are collaborating with Jackson Laboratories to insert the Adult Mouse Anatomy into NCI Thesaurus and to map it to the human vocabulary through appropriate "associations". (See: http://cabig.nci.nih.gov )

Accessibility:

The NCI Thesaurus content is provided under an open content license. See ftp://ftp1.nci.nih.gov/pub/cacore/EVS/ThesaurusTermsofUse.htm. Ontylog XML, OWL and flat file formats are available for download through the NCICB download site. See http://ncicb.nci.nih.gov/download/index.jsp. The content can be browsed through the NCI Terminology server at http://nciterms.nci.nih.gov

The caCORE API provides access to the NCI Thesaurus and other vocabularies provided by the NCI Enterprise Vocabulary Services. See http://ncicb.nci.nih.gov/core/caBIO or contact us for further information.

The OpenGALEN ontology of human anatomy contains some ten thousand anatomical concepts covering all organs, bones, nerves, arteries, veins, ligaments, bursae, muscles, glands etc. and their subanatomy, equivalent to slightly more detail than would be found in an undergraduate medical anatomy textbook. Extensive though not complete or quality checked partOf, branchOf, serves, connects and contains relationships exist between structures, together with inverses of these relationships. Additional relationships denote laterality, or the capacity to have laterality (pairing), as well as links to selector values such as medial / lateral or superior/inferior. Note that the partOf relationship is in fact further subdivided into a number of different flavours of partonomy: layerOf, segmentOf,
pouchOf, irregularPieceOf, componentOf, surfaceDivisionOf and functionalComponentOf.

The model also contains functionOf relations between anatomical structures and the physiological processes they perform, though this is less well populated.

The content has been used to construct and automatically classify separate ontologies of surgical procedures, of pathology and of prescribable drugs.

Syntax: The source files are in GRAIL notation (an early description logic syntax and formalism). A fully open source knowledge management environment exists (OpenKNoMe) to read and edit these, and a closed source but free demo environment is also available. No suitable open source compliant DL engine exists.

Orthogonality: The purpose of the OpenGALEN technology is specifically to provide a means to manage and integrate orthogonal ontologies, so the OpenGALEN anatomy ontology deliberately includes a number of orthogonal axes (including function) to illustrate this.

Defintions: The nature of a Description Logic ontology is to express definitions of the terms within the ontology itself, and to use those formal definitions to infer additional relationships. The description is the concept, and the concept is the description.

Licensing: The content of the full OpenGALEN ontology, and the anatomy subontology, are both open source under the GALEN Open Source License (GOSL). Source files and license are available from http://www.opengalen.org