A wide variety of ontologies, controlled vocabularies, and other terminological artifacts relevant to the biological or medical domains are available through open access portals such as the Ontology Lookup Service (OLS) , and the number of such artifacts is growing rapidly. One of the goals of the Open Biomedical Ontologies (OBO) Foundry initiative  is to facilitate integration among these diverse ontologies. However, such integration demands considerable effort and differences in format and style can only add obstacles to the execution of this task . The heterogeneity within the set of existing ontologies derives from the use of diverse ontology engineering methodologies and is manifest in the adoption by different communities of Description Logic, Common Logic, or other formalisms. The spectrum of syntaxes used to express these formalisms, such as the Web Ontology Language (OWL) or the OBO format, and the commitment of individual communities to conceptualist or realism-based philosophical approaches are also contributing factors.
Here we focus on issues of nomenclature , and specifically on the naming conventions used for labeling classes in ontologies, which are an additional contributing factor to the problem of heterogeneity. Even in this relatively straightforward area, no conventions have achieved broad acceptance (see survey section below).
The lack of naming conventions or their inconsistent usage can impair readability and navigation when viewing ontology class hierarchies. We believe that clear and explicit naming becomes of even greater importance when interlinking ontologies (for example via owl:import, obo dbxref and other referencing and mapping statements , or when ontology engineers need to collaborate with external groups to align their ontologies and to ensure effective maintenance of modularity).
While other sources of diversity are tremendously complex and challenging, it is our belief that establishing a set of naming conventions for the OBO Foundry is a tractable goal, particularly if those conventions are based on lessons drawn from pooled practical experience and targeted surveying.
There is of course no shortage of initiatives for the development of specifications and standards tackling naming [6–9]. However, where naming conventions have been developed, widespread application has been hampered by several factors, most notably domain specificity, document inaccessibility and format dependency. A comprehensive survey of existing naming convention documents can be found at the dedicated OBO Foundry naming conventions website .
One significant obstacle to common adoption is that many of the proposed conventions are domain-specific and not generally extendible to other fields; for example, the Human Genome Organization (HUGO) nomenclature  is restricted to gene names. Other conventions refer only to entities occurring within programming languages  or to the naming of natural language documents .
A second obstacle relates to poor documentation. A naming convention whose documentation is unclear, or is dispersed in multiple documents or document sections, artificially constrains its own chances of acceptance. This is the case with the BioPAX manual , which is in addition overly tool-centric in that it addresses only Protégé-OWL issues. Another deficiency is the commercial or semi-proprietary nature of conventions such as the International Organization for Standardization (ISO) standards . Many of these proposed conventions also impair access through information overload, there being around forty ISO documents addressing naming issues alone. Other naming conventions are described only implicitly and via unintuitive search attributes, or are not available on-line, making access difficult.
Format and implementation dependency
Sometimes only certain naming issues are tackled by a naming convention – usually those most germane to a particular format. The Gene Ontology (GO) Editorial Style Guide  for example, is of limited coverage and applicability, as it is embedded in an OBO-format specific document. The ANSI/ISO Z39.19-2005 Standard  is applicable only to terms organized in an is-a hierarchy without relations and therefore lacks proper conventions for representing ontological classes and properties in semantically complex ontologies.
In the case of the Ontology Engineering and Patterns Task Force of the Semantic Web Best Practices and Deployment working group , the guidelines are restricted to the OWL format and are dispersed throughout many documents and document sections.
To overcome this diversity and fragmentation members of the OBO Foundry and of the Metabolomics Standards Initiative (MSI) ontology working group  have set up an infrastructure group that is attempting to:
collect, review and compare existing naming conventions
distill universally valid conventions that can be implemented in both the OWL and OBO formats, and conceivably also in other formats
engage in discussion with other groups concerned with nomenclature standardization in order to establish a forum for coordinated advance
create a single common guideline document to serve as a common resource for the OBO Foundry and associated initiatives.
In this communication we present the preliminary results of a survey of the naming conventions applied by ontology groups listed under the OBO Foundry, together with an initial set of what we believe are robust conventions for formulation of terms in ontologies and a list of open issues that need to be resolved in the future.