Our proposal is to develop an ontology of qualities (i.e. distinguishing characteristics such as: long, short, increased, decreased, red, blue, and so on) and use it, in conjunction with ontologies of particular anatomies and biological processes, to describe phenotypic data from zebrafish, fruit fly, mouse, and human rigorously. This work will provide one of the integral components necessary for integrating phenotypic descriptions semantically with other aspects of biomedical knowledge. Currently these descriptions are recorded either as free text or using a terminology that is idiosyncratic to a single organism or research project. We have chosen these particular data sets because of the preliminary work that has already been carried out on these data, because they provide a wide spectrum of descriptive situations that will fully exercise the ontology, and particularly because, with the human data, we will be able to explore how unifying these data sets can enable translational research. We will also work with other database resources, such as BIRN [BIRN], WormBase [WormBase], dictyBase [dictyBase], and PhenoScape [PhenoScape], who are also adopting the approach we advocate (see letters of support). We will contribute PATO and the other ontologies that we develop to the OBO Foundry [OBO], and we will deposit our phenotypic annotations into the OBD [OBD] database of the National Center for Biomedical Ontology [NCBO] to facilitate comparisons between these and other data. The Quality Ontology [PATO] will provide one essential part of the unifying framework needed to surmount the difficult problem of integrating phenotypic data sets.
The underlying principle of this proposal is that the comparative approach can be used to further our understanding of the genetic and molecular bases of human diseases. Experiments that study the phenotypic consequences of mutations in non-human organisms are justified by the fact that they provide valuable models for a better understanding of human diseases.
Showing the most recent 10 out of 32 publications