It seems fitting that a major theme from an IT life science conference held on the island of Manhattan involved dealing with integrating islands of research.

At the Executive IT Life Science Forum held last month in New York, much of the talk focused on the increasing problem of how to share information and knowledge to improve collaboration within a company.

Many of the speakers at the conference noted that research today involves individuals and small groups that are so specialised in their research that islands, or silos, of information and knowledge are created within a life science company.

The IT challenge is to find a way to bring these pieces together.

"We must overcome information and knowledge silos," said Tim Clark, vice president of informatics at Millennium Pharmaceuticals. To do this, "we need fully interoperable infrastructures so knowledge can be shared across the R&D pipeline."

Clark serves as chair of the industry group Interoperable Informatics Infrastructure Organisation (I3C).

Clark noted such interoperable infrastructures are needed because "you want people to be mutually aware of developments across disciplines and the [enterprise]."

The problem of sharing information within a company is exacerbated when companies merge, quite common in the life science market today. "Mergers are seen as giving companies economies of scale," said Clark. "I would argue that's true for marketing and sales, but not for R&D."

When a merger takes place, the silo mentality of conducting research means "you have an increased chance that two sets of researchers can duplicate their efforts," Clark said.

Others concurred with Clark. In his talk, Thomas Thuene, manager of pharmaceutical R&D at Accenture, noted that the traditional method of conducting research -- a scientist thinks of an idea, does an experiment to test his or her hypothesis, and then writes down the results on paper -- simply adds to the challenge of sharing information. "Data sources are disparate and too hard to search," said Thuene.

And he sees this problem getting more difficult in the future. In the coming years, "about 40 to 45 per cent of research dollars will be spent outside the organization," Thuene said, as more work is outsourced to contract research organisations.

Many echoed the belief that for drug discovery to improve, interoperable systems and applications that allow for data sharing are necessary. "You need an integrated architecture to bring together all the disparate data -- not [necessarily] physically, but logically," said Sia Zadeh, group general manager of life science division at Sun Microsystems.

Unfortunately, there is no one best way to tackle every life science company's knowledge challenges. Many of the speakers at the conference, however, noted that the two key areas were knowledge management systems, to help researchers interact and share information, and interoperability, so data from all the varied experimental disciplines may more easily be shared and integrated into visualisation, data mining, and knowledge management systems.