Domain concept-based queries for cancer research data sources

Biomedical scientists generate, access, validate and interpret multiple distributed and heterogeneous data sets. Semantic annotations for these data sets are paramount for exchanging and using the data, and take the form of concepts from a domain ontology. ONIX is a platform that facilitates the access to cancer research data resources and one of its goals is to interoperate with caGrid — a grid computing infrastructure for data sharing. In this paper, we present the ONIX approach to building a semantic layer with support for concept-based queries, which exploit semantic annotations of resources, focusing on caGrid resources. The main contributions of this work are: the automatic generation of OWL ontologies from resources' metadata; concept-based query construction and validation; rewriting and translation from concept-based queries to the caGrid query language.