The University is also looking forward to an emerging market in analyzing experimental data. Interest in this area is growing rapidly with the spread of new techniques such as gene sequencing that shed new light on the way cells work. In collaboration with the University of Nijmegen and other Dutch scientific organizations, the University is working on a development called the Virtual Laboratory. The aim is to provide researchers with a new way to collaborate, combining data from multiple sources in order to reveal valuable new information. By mid-year, the University plans to demonstrate proof-of-concept software showing the basic capabilities of the Virtual Laboratory .

Users of the Virtual Laboratory will be able to create experiments that use several different experimental techniques and types of equipment, involving research facilities at several locations. "People who are not in the same place will be able to collaborate on the same problem," Hertzberger says. For harnessing the necessary compute power they will use the emerging Grid technology, a set of services for exploiting multiple geographically dispersed computing resources.

For example, researchers at one University may obtain genetic information from a DNA sequencing experiment. At another research lab, people might gather data about specific cell proteins, using mass spectrometers. Using the Virtual Laboratory, applications could be created that combine and analyze information from both sources, to help predict the ways that the genetic code affects the metabolic pathways and the compounds produced by the cell. This is of great interest to pharmaceutical and biotech firms as well as the academic community, Hertzberger notes.

The plan is that the Virtual Lab will be offered via a portal as a hosted Web-based service. Researchers accessing the portal will get access to a software framework and components that can be pieced together to create experiments and analytical applications they need for their research. The idea is to use the system for a variety of fields, including physics and biomedical applications.

It could be a sizable task just keeping track of the available resources, huge amounts of experimental data and analysis that has already been performed, so that researchers can quickly find what they need. That's where Matisse comes in, acting as an intelligent catalog that stores pointers to the raw data, plus descriptions and analytical information about the data, Hertzberger says. Matisse's flexibility and indexing capabilities are well suited to an environment where large amounts of data are being created, and scientists may want to search the data and combine it in many different ways.

In that sense, Matisse's role is similar to the way it is used in UvA's applications for a completely different domain -- the music industry. It is clear that Matisse's strengths -- such as flexibility and its ability to handle complex objects such as multimedia data -- are proving a good fit for more than a few emerging-market applications.