Biosof, a three-year-old New York city-based bioinformatics software shop, said this week that it has secured the rights to commercialize a suite of software tools developed at Burkhard Rost’s lab at Columbia University Medical Center’s Department of Biochemistry and Molecular Biophysics.
The agreement covers several software packages, including PredictProtein, a web-based protein structure prediction server that has been freely available since 1992. Biosof will offer a “premium” version of the software that will offer a quicker turnaround time for analysis than the current version, according to Donna See, a licensing officer with Columbia’s science and technology ventures department.
See said that Columbia has had a number of requests from industry for a commercial version of PredictProtein. The software will remain free for academics, but Biosof will offer customization, software upgrades, and ongoing support, she said.
“The novelty is the introduction of a ‘premium service.’ Users of that service pay a minimal fee to enable a faster turnaround of requests and get more substantial support,” See added.
Guy Yachdav, founder of Biosof, said that the company will offer consultancy services in tandem with the offering. He could not disclose further terms of the agreement.
PredictProtein will join the company’s current software lineup, which includes PROFphd, a suite of protein-structure and function-analysis tools; AGAPE, a protein comparison tool that uses 1D structure; LOCTree, software for predicting sub-cellular localization; and Genetegrate, a semantic data-integration platform. All of the software was developed by Rost.
Under the terms of the agreement with Columbia, Biosof will also distribute CONBlast, a Blast search algorithm; PiNAT, a systems biology modeler; and UniqueProt, a solution for creating unbiased protein sequence data sets.
Biosof currently employs five people. Yachdav told BioInform that the company has so far focused on selling PROFphd to mid-sized pharmaceutical firms and some biotech companies.
“This was mostly client driven, then actually [involved] us going out and doing marketing. But now we’ve shifted gears and actually create our solutions with the strategy of selling to the larger pharmaceuticals and R&D [labs] in the country.”
PredictProtein is a “strategic investment” for Biosof “in that it will get a foothold in the industries that we are targeting,” Yachdav said. “Once we get that foothold, we will offer our consulting services.”
Columbia’s Rost told BioInform that one reason his group decided to commercialize PredictProtein was the rapid growth of public protein databases, which has bogged down the freely available service. As these databases have doubled in size, “the issue became twice as important,” Rost said. “The turnaround time used to be three days and now it’s six days. We clearly needed to find a way to get the research done.”
“The growth of these databases has been outpacing the growth in the power of the computing hardware available to realize the searches.”
See reiterated Rost’s comment. “PredictProtein searches through databases of proteins. The growth of these databases has been outpacing the growth in the power of the computing hardware available to realize the searches.”
The result, she said, is that it takes increasing amounts of time to search through these databases in order to identify evolutionarily related proteins. “That means longer wait times for users at a moment in the development of modern life at which we all increasingly expect push-button service.”
She added that with more and more protein information entering public databases from metagenomics projects and other efforts, “we felt that the time has come to take the challenge and attempt at speeding up through [a vendor’s] financial backing.”
Further, she said Columbia believes that Biosof “will be able to build upon the current success of the software by expanding access to and improving functionality of the technology, to keep up with the needs of the growing user community. A commercial vehicle is better suited to address these market needs and requirements” than an academic model.
Rost said that any future commercialization plans for his software will be decided “on a case-by-case basis.”