VBO - A Tool for Bridging Vertebrate Anatomy Ontologies

Anatomy is a central concept to the biomedical sciences. For multicellular animals the differentiation of cells into different structures and cell types, their arrangement in space - their Anatomy - defines the organism. Biologists need formal, unambiguous ways to describe Anatomy, and whilst the familiar Latin terms to describe our arms, legs, fingers and blood vessels for example have stood us in good stead for centuries, we are now in a position where the amount of data being collected on, for example, where genes are expressed, needs to be annotated in ways that a computer can understand. An approach to this, now widely used, is to make an 'ontology' which is a tree-like arrangement of anatomical terms in a common-sense hierarchy. The 'big toe' is a part of the 'foot' which is a part of the 'leg' etc. Such ontologies have been made for several species such as human and mouse. However, as there are differences in anatomy between species such single species anatomy ontologies are not re-usable, even between closely related species. This project addresses this problem by developing a system whereby we can link the anatomical descriptions of structures in humans, mice, fish and birds through the common evolutionary ancestor of the structure. Thus we can link the bird's wing, the mouse's paw and the fish's fin through their common evolutionary ancestral structure. Molecular biology tells us that this is a true linkage and that genes responsible for the development of all of these structures have a great deal in common and in many cases are orthologs. By developing this computable description of detailed cross-species anatomy we will enable scientists currently collecting large amounts of data on gene expression, abnormal morphology and biological processes in different tissues to compare their data computationally with that being collected by other scientists working in different vertebrate species. This will allow the discovery of new information about the function of genes and the interaction of gene products in the organism and ultimately inform us about the way that humans and other species are similar, or different. Computational comparison is needed as modern technologies allow scientists to compare gene expression - where and when a gene is active, for thousands of genes at a time.

The VBO ( Vertebrate anatomy Bridging Ontology) project will deliver a vertebrate cross-species anatomy ontology developed using a novel approach; that of the Most Recent Common evolutionary Ancestor (MRCA). This cross-species anatomy ontology will fill a critical gap in data annotation for systems biology, functional genomics and high throughput data gathering by allowing computational cross-species integration of data between the key model vertebrate organisms. Cross-species integration of such data along the anatomical axis is currently not possible due to the lack of a public domain cross-species anatomy ontology which is implementable by existing data resources. This is a serious problem as, for example, the wealth of data from comparative genomics is not aligned anatomically in any meaningful way and cross-species gene expression data cannot be accurately queried by anatomical location. The novel approach used for its construction will allow for application to species without explicitly derived anatomy ontologies and extension to new organisms. Development of VBO will allow leveraging of the data in the public domain and provide added value through integration and comparative analysis of datasets derived from multiple species. In addition, the new paradigm it exploits will support generation of new computational approaches to data integration by the knowledge management community, an area where ontology development is burgeoning. This is a community project and will gather input from a range of potential users and domain experts. The Project brings together biologists and computer scientists working with the biology and informatics of the major vertebrate model organisms, and builds on a UK strength in this area. It will engage with the wider user community through two workshop meetings and an extensive web presence. All project resources will be made publicly available via Open BioMedical Ontologies (OBO) and relevant databases.