The biomedical ontology community is producing ontologies which represent biological knowledge and with a bias towards a realist perspective due to the use of the upper level ontology BFO [1]. While these ontologies are useful representations of the knowledge space they are not always applicable in the context of typical biomedical use cases. In this paper we describe our use of reference ontologies in the development of the Experimental Factor Ontology (EFO) and its application to public gene expression data. Finally, we present guidelines for rapid development and deployment of an application ontology.