Community Driven Ontology Development

Frank Gibson and James Malone^§

^§European Bioinformatics Institute, Cambridge, CB5 8LW, UK

Community driven ontology development is the process of collaboratively building an ontology which represent the understanding of a particular community or domain area. Within the biological domain, collaboration and community involvement is common place. As an ontology can be interpreted as a “shared understanding” of a particular domain, collaboration and community involvement should be maximised within the life-cycle of an ontology.

Background
Integrating biological knowledge within an ontological framework produces what is referred to as a bio-ontology, where a shared understanding of biology is represented in a computationally amenable form.  The study of the biological sciences is a global effort of researchers and institutions each specialising in, or across particular niches to further our understanding of biology. As the experts in a particular biological field are rarely physically co-located Bio-ontology development as a result is highly distributed, forming what can be thought of as virtual organisations in which experts with different but complementary skills collaborate in building an ontology for a specific purpose.

Typically bio-ontology development is dynamic where different domain experts join and leave the network at any time and decide on the scope of their contribution to the joint effort. In addition, biological ontologies continue to evolve, even after the initial development drive. The continued evolution reflects the advancement of scientific knowledge discovery. New classes, properties, and instances may be added at any time, and new uses or extended scope for the ontology may be identified .

The diversity of the life-science domain results in a multitude of application domains for ontology development and therefore produces numerous ontologies for biology. However, with the diversity, there is equal homology. Typically, the same experimental equipment and reagents can be used to study different aspects of biology. For example, a mass spectrometer can be used to determine the elemental composition of a molecule in a chemistry based experiment, and to determine the chemical structure of peptides in a proteomics investigation.  The multiple application of equipment, reagents and organisms in the study of biology and the potential to be represented multiple times across different bio-ontologies, with slightly different defintions. This potential proliferation or duplication of terms, could undermine the ethos of ontology development – to produce a shared understanding.

Examples

OBI
The Ontology of Biomedical Investigations (OBI) aims to produce an ontology which represents the common components of life-science experimentation, such as equipment, materials and protocols. The developer community of OBI is currently affiliated with 18 diverse biomedical communities, ranging from functional genomics to crop science to neuroscience. In addition to having a diverse community of expertise, the OBI developers work in a virtual organisation encompassing multiple countries and time zones.

The OBO Foundry
In an attempt to address the issue of bio-ontology proliferation and potential overlap the The Open Biomedical Ontology (OBO) Foundry was formed (Smith et al, 2007). The OBO Foundry describes itself as “a collaborative experiment involving developers of science-based ontologies who are establishing a set of principles for ontology development with the goal of creating a suite of orthogonal interoperable reference ontologies in the biomedical domain.”  The role of the OBO Foundry is twofold. In one role The OBO Foundry acts as a registry to collect public domain ontologies that, by design and revision, are developed by and available to the biomedical community, fostering information sharing and data interpretation. This section is called the OBO library. (As of 23rd of January 2010 there are 111 candidate ontologies at the OBO Foundry, representing knowledge domains ranging from Amphibian gross anatomy, infectious diseases to scientific experimentation. In addition to providing a library of bio-ontologies, the OBO Foundry was formed to reduce ontology overlap and ensure bio-ontology orthogonality. Initial steps at achieving this aim have produced a set of design {http://www.obofoundry.org/crit.shtml} to which domain ontologies should adhere, such as, openness, a shared syntax and class definitions. The OBO Foundry proposes a review process by which ontologies may become ‘certified’ as meeting the OBO Foundry criteria (e.g. orthognality, shared namespace. At present no OBO Foundry ontology has been awarded this certified status, although several have been proposed as “Candidate ontologies” which are ready to be reviewed. Criticims of the OBO Foundry centre around the lack of suggested community orientated engineering methodology , method or technique by which these principles can be met (see Hull and Gibson).

Authors: Frank Gibson and James Malone

Acknowledgements

This paper is an open access work distributed under the terms of the Creative Commons Attribution License 2.5 (http://creativecommons.org/licenses/by/2.5/), which permits unrestricted use, distribution, and reproduction in any medium, provided that the original author and source are attributed.

The paper and its publication environment form part of the work of the Ontogenesis Network, supported by EPSRC grant EP/E021352/1.