Ten Simple Rules for Selecting a Bio-ontology

Biologists and bioinformaticians now look to ontologies or software that uses ontologies as a means of standardising the way data are described, queried, and interpreted. Ontologies can be used for the annotation and curation of experimental datasets and, in data sharing, both within and beyond the confines of individual labs, organizations, and communities. Bio-ontologies are also commonly used in methods of analysis, particularly in gene set enrichment analysis [1], using ontologies such as the Gene Ontology. With modern high-throughput data-generation technologies, there is now, more than ever, a need to integrate data from these and other sources, and there is a concomitant need for ontologies—raising the question of how to choose a bio-ontology. Over the past decade, a community has grown up around the success of efforts to harmonise the semantic description of biological entities, with ontologies exemplified in the emergence of the Open Biological and Biomedical Ontologies (OBO) Foundry [2]. These efforts were first led by the aforementioned Gene Ontology [3] and have expanded to ontologies that describe a significant range of the primary areas of biology and its science. Exploring bio-ontologies through browsers such as the Ontology Lookup Service [4] at the European Bioinformatics Institute and BioPortal [5] at the National Center for Biomedical Ontology (NCBO)—whose existence is itself a measure of the community size—shows there are over 400 ontologies containing, collectively, over 5 million classes (by classes, we mean ontological terms together with their associated descriptions and synonyms). These ontologies cover areas such as diseases [6], phenotypes [7], anatomy [8], experimental conditions [9,10], cell types [11], and bioinformatics software [12]. There are now many ontologies from which to choose, but which ontology should be chosen? In order to answer this question, we present ten simple rules that should help to guide the choice of a bio-ontology. The rules are designed to be useful for those wishing consume a bio-ontology. Users of bio-ontologies are varied in their profile and include data curators, application developers, and, of course, ontology developers who may be consuming part of an ontology in their own work.