While centralized respositories such as Genbank have proven to be valuable resources, they require significant resources to maintain and often don't reflect the needs of individual laboratories. This is especially true of gene expression data where the contextual data is equally as important as the numerical expression values. Many laboratories will want to organize their data in local data management systems, and each system will likely vary in the amount and type of contextual data that it records. One issue is that centralized databases will likely be unable to store all of the contextual information associated with every experiment (because their data model only supports a common denominator of all experiment types) or unwilling to (because of the resources required to store that level of information).

The GeneX project (genex.sf.net) has begun exploring Peer-to-Peer mechanism for publishing scientific data; help peers find what data is available, enable peers to query the data or download data to their local machines. GeneX includes a framework based upon the successful model of the Distributed Annotation System project (DAS - biodas.org), we call G2G.

G2G defines a distributed network of cooperating gene expression databases that use a common basic data model, standards for minimal information about microarray data, and standardized exchange formats (defined by MAGE - mged.sf.net). Each peer is responsible for the quality of data they control which distributes the cost and responsibility across the community.