Computational phenotypic analysis of human hereditary disease

• Applicant: Professor Dr. Peter Nicholas Robinson
• Subject Area: Human Genetics
• Term: from 2008 to 2015
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 103545043

We have developed the Human Phenotype Ontology (HPO) as a tool for annotating and analyzing the manifestations of hereditary and other diseases in humans. We have developed a number of algorithms for using the ontology and annotations to it for clinical differential diagnosis and computational analysis of the phenome. The HPO has begun to be adopted by international groups involved in genetics research including the Sanger Centre, Orphanet, and the International Standards for Cytogenomic Arrays (ISCA) consortium. This proposal involves the development of novel computational algorithms for phenotype ontologies. Bayesian network algorithms will be adapted for bio-ontologies in such as way as to compensate for the statistical dependencies typical of hierarchically structured ontologies. The algorithms will be used to improve programs used for helping geneticists and others in the differential diagnosis of rare diseases. Tools of the semantic web will be used to link the HPO to data representing anatomy, biochemistry, pathology etc. in order to provide a logical map between human and model organism phenotype data. This will immediately be useful for mapping data from the International Knockout Mouse Consortium and related efforts.

• DFG Programme: Research Grants