Cyberinfrastructure for Biology
The BIRN Project

The Biomedical Informatics Resarch Network, funded by the NIH’s National Center for Research Resources since 2001, supports a growing number of collaborative projects that involve more than 30 universities and 40 research sites. These collaborative groups are primarily centered around the storage and analysis of neuroanatomical, clinical, genomic and behavioral data in humans and animal models. Much of the data that BIRN works with is structural and functional imaging data, including magnetic resonance (MRI) and functional MRI studies,, which require large amounts of storage space and processing power. A distinguishing characteristic of BIRN is that it has developed a robust software installation and deployment system that allows for research groups to easily deploy a local BIRN end-point, called a “BIRN” rack. Once a research center has installed a rack, which costs roughly $20,000 US, it can host data and contribute compute resources to the BIRN grid. Registered users can access shared BIRN datasets and computational resources from any internet capable location via a web portal, that provides a collaborative environment for the research scientist, or through a collection of diverse data mangement, analysis and visualization applications. BIRN also provides free access to published datasets to interested researchers via a data archive called the BIRN Data Repository.

BIRN uses the Globus Toolkit to attach compute and computational resources to the grid and to authenticate authorized users. Semantic integration is achieved by BIRNLex, an ontology that covers multiple aspects of neuroanatomy, species, behavioral and cognitive processes, subject information, experimental practice and design, and provenance information. A large portion of BIRNLex is based on shared community ontologies, such as NeuroNames, the NCBI Taxonomy, and OBO Foundry ontologies such as the Ontology for Biomedical Investigations (OBI) and PATO.