International Mammalian Genome Society

17th International Mouse Genome Conference

9-12 November 2003, Braunschweig, Germany

Plenary Presentations * Oral Presentations *
Poster Presentations: Behavioural Genetics and Genomics * Development and Stem Cells * Functional Genome Analysis * Mouse Models of Human Disease * Mouse System Biology Bioinformatics * Multigenic and Multifactorial Trait Analysis * Nutrition and Metabolic Disease * Phenotyping Methods Imaging * The Genetics and Genomics of Infectious Disease *
Verne Chapman Memorial Lecture * Table of Contents * Sponsor/Exhibitor List * Awards * Photographs

---

POSTER 138 - KNOWLEDGE DISCOVERY FROM MOUSE MUTAGENESIS PROJECTS USING ONTOLOGIES

Gkoutos G
Bioinformatics Group, MRC Mammalian Genetics Unit, Harwell, Oxfordshire, U.K.

Co-Authors: 1) Green E, 1) Blake A, 1) Simon M, 1) Greenaway S, 1) Mallon A-M, 1) Hancock J M, 2) Davidson D.
Institutions: 1) Bioinformatics Group, MRC Mammalian Genetics Unit, Harwell, Oxfordshire, U.K., 2) MRC Human Genetics Unit, Edinburgh, U.K.

With the completion of the mouse genome sequence in sight, emphasis is shifting to characterisation of gene function in this organism, which is the major experimental model for human disease. A number of large centres worldwide are investigating gene function using a variety of mutagenesis studies combined with a battery of assays designed to identify phenotypic changes resulting from induced mutations. The types of change assayed for range from blood biochemistry to behaviour. This poses the intriguing problem of interlinking data of these different types, often held in different databases, with more conventional genomic (e.g. sequences) and functional genomic (e.g. microarray) data. This should allow the discovery of relationships within the data that may shed new light on gene function and its relationship to phenotype.

Ontologies present an increasingly popular approach to accessing expert knowledge and using it to annotate and mine data. Linking phenotype and genomic data together will involve using a combination of ontologies representing a variety of different domains of knowledge. Some of these ontologies are already well developed, for example the GO ontologies and, at the organism level, ontologies for developmental and adult anatomy of the mouse. Other domains are less well catered for, including the complex domains of behaviour and pathology. As part of a wider community effort to construct phenotype ontologies along standard lines, and as part of the EUMORPHIA consortium, we are participating in the development of aspects of a mouse phenotype ontology and addressing theoretical problems of reasoning with ontologies.