Building better animal models of disease through integration of multiple disease phenotypes

Abstract

Animal models of disease are used to test hypotheses of disease mechanisms as well as for therapeutic development. Effective animal models must sufficiently reflect human disease and should be classified carefully as disease models. Integration of genomic and phenotypic data provides the foundation to study the genetic and environmental influences on the risk, onset and progression of disease. PhenoMiner, at the Rat Genome Database (http://rgd.mcw.edu), is an interactive data‐mining tool to compare clinical measurements used as indices of disease across multiple rat strains, measured under a variety of conditions. Data are integrated through the use of four ontologies (rat strain, clinical measurement, measurement method, experimental conditions). PhenoMiner provides a solid foundation to distinguish phenotypes by strains and strains by phenotypes. Currently more than 42,000 records can be viewed graphically or in tabular format, allowing easy comparison across strains. Graphical summaries (bar charts and heat maps) provide a rapid method to identify strains with abnormal phenotypes or groups of phenotypes that correspond to disease traits. In addition to providing phenotype driven model identification, PhenoMiner generates standard ranges for clinical measurements under control conditions to be used to identify strains with potential disease traits.Supported by R01HL094271 and R01HL064541.