VOLUME 24 NUMBER 1 JANUARY 2006 NATURE BIOTECHNOLOGY manner that affords straightforward, reliable and increasingly automated access to both data and the methods by which they were generated, and is helping to meet the timecritical demands of current research. The MO is one of the early examples of an ontology that was developed by a community in that spirit, specifically to describe microarray experiments. In their commentary, Soldatova and King show no awareness of the challenges posed by rapidly evolving, high-throughput genomic technologies that are generating data at an exponentially increasing rate. Controlled terminology as provided through an ontology was urgently required for the description of microarray experiments; biologists did not expect emerging ontologies to be ideally engineered, and we would be the first to state that the MO contains compromises. In the absence of such controlled terminology, much of the data being generated today would not be available in a form readily amenable to comparative analysis, either now or in the future. Fundamental inaccuracies in this critique appear to reflect a lack of understanding of the MO’s nature. For example, the authors state that the MO “has been promoted as an international standard.” The MAGE object model is an international standard for exchange of microarray data. The MO was developed to provide terminology required to support the annotation of microarray experiments in areas where the MAGE Object Model does not provide descriptors. Furthermore, the authors state that the Minimum Information About a Microarray Experiment (or MIAME) specification is at the root of some of the design compromises in the MO. This is a puzzling assertion; surely they meant to refer to the MAGE Object Model, from which some of the classes in the MO were created. The authors suggest rebuilding the MO. But the benefits of such refactoring need to be weighed against the necessity of maintaining a stable real-world resource that is widely applied in software implementations. This is in fact the reason why there is a ‘core’ component to the MO deliberately held stable—a design decision that seems to be misunderstood by the authors. Additionally, the Functional Genomics Ontology (or FuGO; http:// fugo.sf.net) project is now underway to collaboratively develop controlled terminology for transcriptomics, proteomics and metabolomics, as used in a range of biologically defined domains. The decomposition and reuse of the MO will be part of the process of building FuGO, allowing the MO itself to remain stable, continuing to serve its established user base. The authors suggest the building of ontologies that are (i) purpose independent and (ii) compliant with a standard upper ontology. These ‘key rules for bio-ontology development’ represent unrealistic expectations of the current state of play as the adoption of an upper ontology is predicated on its acceptance by user communities who must decide whether or not an upper ontology is fit for its intended purpose. There are many ‘standard’ upper level ontologies. Such ontologies are themselves subject to major philosophical debate. In addition, a valid demonstration of their utility in ontology building is by no means universally accepted, even within the computer science community. Finally, we address perhaps our greatest concern with respect to this commentary: the article centers on the “isolation of bio-ontologies from the larger world of ontologies” and asserts the need for interaction between MO developers and other ontologists. It is therefore ironic, and of notable concern, that the authors of this commentary have never communicated their analysis, or suggestions for improvements to any of the extended family of MO developers, that includes ontologists in an advisory capacity. This sort of collaborative resource development is not a competitive exercise. We have ongoing interactions with a number of leading ontologists, and we encourage other researchers to contact us (http://mged. sourceforge.net/ontologies/) to share their views and help ensure that good ontologies, which benefit the bioscience community, are developed as effectively and efficiently as possible.