Interoperability across neuroscience databases.

Abstract

Data interoperability between well-defined domains is currently performed by leveraging Web services. In the biosciences, more specifically in neuroscience, robust data interoperability is more difficult to achieve due to data heterogeneity, continuous domain changes, and the constant creation of new semantic data models (Nadkarni et al., J Am Med Inform Assoc 6, 478-93, 1999; Miller et al., J Am Med Inform Assoc 8, 34-48, 2001; Gardner et al., J Am Med Inform Assoc 8, 17-33, 2001). Data heterogeneity in neurosciences is primarily due to its multidisciplinary nature. This results in a compelling need to integrate all available neuroscience information to improve our understanding of the brain. Researchers associated with neuroscience initiatives such as the human brain project (HBP) (Koslow and Huerta, Neuroinformatics: An Overview of the Human Brain Project, 1997), the Bioinformatics Research Network (BIRN), and the Neuroinformatics Information Framework (NIF) are exploring mechanisms to allow robust interoperability between these continuously evolving neuroscience databases. To accomplish this goal, it is crucial to orchestrate technologies such as database mediators, metadata repositories, semantic metadata annotations, and ontological services. This chapter introduces the importance of database interoperability in neurosciences. We also describe current data sharing and integration mechanisms in genera. We conclude with data integration in bioscience and present approaches on neuroscience data sharing.