Now that distributed computing technology is gaining acceptance within life science research organizations, it may be time to consider distributed knowledge management. Varro Technologies, a Philadelphia-based startup, is betting that the industry is ready to try.
Varros technology, called VarroGen, is a combination of Oracle and Napster, according to CEO Jay Venkatesan. The system uses proprietary search technology, XML databases, a distributed network environment, and peer-to-peer messaging technology to allow users to share information stored in their desktops across a network.
Currently, the information that exists on individual computers is not being shared with the rest of the organization in a useful system, said Venkatesan. While some organizations employ centralized data warehousing systems, Venkatesan said this approach is limited for a number of reasons: the data copied to the warehouse is only a small portion of that available on individual computers, the data is not updated as frequently as it may be on individual computers, and the search speed is constrained.
In addition, Venkatesan said, You have to do a lot of data massaging to get it into a standardized form that a relational database can use. By contrast, he said, VarroGen offers a much more flexible and extensible data structure.
The system has three components: a messaging component that allows users to interact in a peer-to-peer fashion; a centralized XML database that serves primarily as an indexing function to direct queries; and a client-side application that allows users to manage data and annotation information.
Venkatesan hopes the pharmaceutical industry will view the technology as an efficient way to minimize duplicated research efforts. If a researcher can take advantage of the fact that someone else has already completed work in a particular area, we can easily trim one week to three months from a project, he said. He estimated that a typical pharmaceutical company could save $2,000 a day for each day of redundant work eliminated.
Once hitting upon the file-sharing concept as a useful approach to managing genomic information, Varro considered licensing Napsters technology, but soon realized it wouldnt be able to handle the complexity of genomic data. Once you start talking about searching for sequences and similarity searching as opposed to just simple file name searching, you go up a magnitude of difficulty. And when you start talking about the amount of data, the amount of data types, and the relation between those data types, you go up another order of magnitude, so we decided the only way to do it was to do it ourselves.
Varro recently licensed EcomXMLs EcomTalk software for data transformation and transfer and chose NeoCores XML Information Management System (XMS) as the information platform for the VarroGen system.
An undisclosed pharmaceutical company is now beta testing VarroGen and Varro is in talks for another beta project with another pharma. Venkatesan said the company expects to release the product commercially in the first quarter of 2002.