For this project, we use a non-monotonic logic to integrate anatomy and phenotype ontologies.

We use for our implementation the dlvhex system, that allows the combination of default rules with OWL ontologies.

We use nonmonotonic reasoning to integrate two kinds of ontologies, ontologies of canonical anatomy and phenotype ontologies. For testing purposes, we use the Mammalian Phenotype Ontology and the Mouse Anatomy Ontology. The result is the following OWL file: all.owl.

This OWL file is used in conjunction with the dlvhex system, and the following answer set program: int.dlv. It can be invoked using dlvhex int.dlv. The command dlvhex -f hasSomePart int.dlv filters for parts of some individual. Note that dlvhex must use the OWL and RDF plugin in order for this command to succeed. In addition, the RACER reasoner must be running and DLV must be present.

The complete dataset for the performance evaluation and the file conversions is available for download: nmr.tar.gz