While the minimum data standards describe the types of data elements to be captured, the use of standard vocabularies as values to populate the information about these data elements is also important to support interoperability. In many cases, groups develop term lists (controlled vocabularies) that describe what kinds of words and word phrases should be used to describe the values for a given data element. In the ideal case each term is accompanied by a textual definition that describes what the term means in order to support consistency in term use. However, many bioinformaticians have begun to develop and adopt ontologies that can serve in place of vocabularies for use as these allowed term lists. As with a specific vocabulary, an ontology is a domain-specific dictionary of terms and definitions. But an ontology also captures the semantic relationships between the terms, thus allowing logical inferencing about the entities represented by the ontology and by the data annotated using the ontology’s terms.
The semantic relationships incorporated into the ontology represent universal relations between the classes represented by its terms based on knowledge about the entities described by the terms established previously. An ontology is a representation of universals; it described what is general in reality, not what is particular. Thus, ontologies describe classes of entities whereas databases tend to describe instances of entities.
The Open Biomedical Ontology (OBO) library was established in 2001 as a repository of ontologies developed for use by the biomedical research community (http://sourceforge.net/projects/obo). In some cases, the ontology is composed of a highly focused set of terms to support the data annotation needs of a specific model organism community (e.g. the Plasmodium Life Cycle Ontology). In other cases, the ontology covers a broader set of terms that is intended to provide comprehensive coverage of an entire life science domain (e.g. the Cell Type Ontology). The European Bioinformatics Institute has also developed the Ontology Lookup Service (OLS) that provides a web service interface to query multiple OBO ontologies from a single location with a unified output format (http://www.ebi.ac.uk/ontology-lookup/). Both the BioPortal and the OLS permit users to browse individual ontologies and search for terms across ontologies according to term name and certain associated attributes.