While we launch CMG as a long-term community tool, other consortiums and databases are facing a tough future. In Europe, database funding is often linked to research projects. Europe has nurtured a significant number of databases of global impact, such as Swiss-Prot; and the EBI in Cambridge continues to be a focal point for cutting edge bioinformatics. However, a database can only thrive when it is well maintained with regular data-feeds. It is lamentable that European governmental funding does not usually secure the long-term sustainability of databases and several worthwhile projects are facing an uncertain future as a result (see Nature, 23 June). Indeed, some databases have become commercial as a result; this can constrain access to the data, as was seen all too clearly when the Yeast Protein Database became part of the commercial entity BIOBASE.

While the NIH is the main supporter of the umbrella protein database UniProt (which contains the European databases Swiss-Prot and TrEMBL) and provides funds for a number of key organismal databases such as FlyBase, all is not so well on the other side of the Atlantic. The Canadian government, for example, insists that its funding be matched by independent sources, which has led to the cancellation of funding for a number of large-scale projects (see Science, 24 June). One such project is the BIND database (see August 2004 editorial), which is set to shift activities to Singapore as a result (Science 308, 1529; 2005). Another project is the US-based Alliance for Cell Signaling (AfCS), which faces a sizable reduction for the second term of their NIGMS 'glue grant'. Aside from curtailing the data acquisition part of the project, this has necessitated the move to acquire alternative funding to ensure sustainability of the Molecule Pages database and the Signaling Gateway (see http://www.signaling-gateway.org/).

Evidently databases are proliferating and some redundancy is emerging. However, no more so than with bench research, and some overlap is to be encouraged to ensure projects improve in a competitive environment. The information explosion in the biological sciences necessitates that funding agencies invest as much in the initiation of these database projects as in their long-term future, which is usually a fraction of the initial cost. It is hoped that funding agencies can transcend national boundaries and traditional funding criteria to ensure that the best databases become permanent research tools — only this will allow a full exploitation of the promise of systems biology.