We introduce the BioFuice approach for integrating data from different private and public data sources and ontologies. BioFuice follows a peer-to-peer-like data integration based on bidirectional mappings. Sources and mappings are associated with a domain model to support a semantically meaningful interoperability. BioFuice extends the generic iFuice integration platform which utilizes specific operators for data fusion and workflow-like script programs. BioFuice supports explorative data analysis and query and search capabilities. We outline the integration approach by an illustrating scenario, the architecture of BioFuice and its query interface.