Who says patience is a virtue? The recent collaboration of Ohio State University, Ohio Supercomputing Center, and LabBook was borne largely of the unwillingness of Ohio State researchers to wait for the genome map to make sense.
Though it wasn’t a player in the Human Genome Project, the Ohio State Human Cancer Genetics Program knew that genomic data would be necessary to “put Ohio State on the map as far as cancer research goes,” says Fred Wright, a professor in the department.
Hence the OSU Genome Database, an annotated map that its creator, Bo Yuan, claims is authoritative enough to compete with the products at EBI, NCBI, and Celera. Yuan, head of bioinformatics for the cancer program, says the OSU map is based on the compilation and integration of three independent maps — UniGene, dbEST, and the draft of human genome sequence — which makes the information more accurate than other sources.
Wright explains, “Any redundantly represented ESTs will resolve themselves when you put them on the sequence. Splicing gives you confidence that you’re looking at a real gene as opposed to what you thought was an EST but is actually a contamination.”
Yuan adds that a particular advantage with the OSU data is that “our gene index was based on real transcription rather than on gene prediction.” His team used the Ohio Supercomputer Center’s 132-node Silicon Graphic 1400L cluster to maintain the information.
Still, they needed a better way to access it.
Enter LabBook. Wright says, “We’ve done a lot of informatics work to compile the databases and boil it down to useful information. We were very interested in access to [LabBook’s XML] browser so that the annotation we had could be viewed in a highly visual environment.”
According to Adel Mikhail, vice president of strategic development at LabBook, that’s exactly what Ohio State got. Working in conjunction with cancer genetics scientists, the company developed a hypothesis-driven front end that allows users to query the database with normal descriptive terms. “It allows biologists, not just bioinformaticists, to query the database,” Mikhail says. The information is returned in a “dynamically active rendered graphic” with the annotation, and could include links to literature or commercial annotations as well.
The database will be available exclusively through LabBook in XML, and also will be offered in HTML or an XML-light viewer at no cost to academic users, Mikhail says.
— Meredith Salisbury