The Blueprint Initiative announced this week that in addition to its current offices in Toronto, Canada and Singapore, it will be launching a new office in London that will collaborate with Nature Publishing Group to further expand the Biomolecular Interaction Network Database, or BIND.
BIND is a database of records documenting molecular interactions, including protein interactions as well as interactions between DNA, RNA, ligands, molecular complexes, photons and unclassified biological molecules.
“We’re trying to get to the point where capturing (molecular) interactions from journals is not so much like catching rain in a teacup,” said Christopher Hogue, the principal investigator at Blueprint Initiative, a research program based in the Mount Sinai Hospital in Toronto. “It’s hard to do. The first step is to get the tier one journals involved. Science and Nature have both endorsed our curation process.”
Hogue estimated that at maximum capacity, curators trained to cull interaction data from journal articles, could add about 4,500 new entries per month into the database. In addition, collaboration with other databases further increases the number of molecular interactions collected in BIND.
“In June of 2004 we surpassed adding 80,000 interactions into the BIND database with curation activities underway at the Blueprint Initiative’s headquarters in Canada,” said Hogue. “This important milestone fulfills a promise that we made to our Canadian funding agencies one year ago.”
Endorsement of the database by major journals is an important step in producing a comprehensive molecular interaction database because it could lead to a more automatic process where scientists submit information to the database themselves, Hogue said. Eventually, the system might become like Genbank, where scientists with sequencing results are required to submit their findings to the database.
By encouraging scientists to submit to the database themselves, journals could ease the burden on curators to find and enter data using published articles. A submission system would also make the database more up to date, with molecular interactions being published in the database while the paper is still in the pre-publication process.
Blueprint’s Singapore offices opened in February of this year, and London offices were expected to open by January 2005.
Currently Blueprint employs 25 curators in Toronto and seven in Singapore. With additional funding and equipment from Sun Microsystems and Singapore’s Economic Development Board, the organization was expected to employ up to 40 additional curators.
“We’re very excited and we’re working very diligently to make sure we can accommodate the management,” said Hogue.
In 2001, Blueprint received about $29 million worth of Cana dian funding to develop its project over a period of three years. The organization expected to renew that funding for the next few years, and to receive additional funding from Sun Microsystems and Singapore, Hogue said.
Currently, there are about 80 journals that publish molecular interaction data, Hogue estimated. The two journals that publish the most interaction data are the Journal of Biological Chemistry and EMBO, said Hogue.
According to Hogue, Blueprint staff surveyed over 120 journals over a three month time frame to ascertain the rate of influx of recordable molecular interactions. Interactions published in JBC alone accounted for about a quarter of the entries.
“Our database is distinguished by its depth and breadth. We do all types of molecules and all types of organisms,” said Hogue.
Other molecular databases that currently compete with BIND include the Database of Interacting Proteins (DIP), which is maintained at the University of California Los Angeles and MIPS, a yeast protein interaction database maintained in Munich.
In order for BIND to become the industry standard database for molecular interactions, it is important that the data that is entered be peer reviewed and legitimate, and that the database receive the support of scientific journals.
Hogue said Blueprint is unique because the curation process for the BIND database is published on its website. In addition, all curators go through a training process to make sure that curation of papers is the same, regardless of which curator goes through the scientific papers.
“(The training) controls the process of transforming published material into database records,” said Hogue. “These people are hands over heart making sure that every record doesn’t have the personality of the human that made the record.”
—TSL