By Malorye Branca
January 15, 2003 | Don’t throw out your computers yet. Bioinformatics is still the linchpin of genomics.
True, the sector has experienced a steep decline, beginning in winter 2001, when Incyte Genomics and Celera Genomics refocused on drug discovery. By t fall 2002, bioinformatics was in serious trouble. Companies were going bankrupt, exiting the field, and being acquired at bargain prices.
One problem was that many “bubble era” startups tried to sell big-ticket items that lacked interoperability. “They wanted to get a lock on the customer base,” says Mick Savage, an independent bioinformatics consultant. Most clients didn’t like that idea and just kept building and buying various products.
Overfinancing didn’t help, either. The field became crowded, which fostered feature creep.
“Instead of doing novel things, you’re adding little tiny features because your competitor added that, and so you have to do it so the sales guys can close,” Savage says.
Clients weren’t getting enough value, but the blame doesn’t rest entirely on the vendors. Biologists were not being trained to handle high-throughput tools, so many were generating flawed data.
The result was overall disappointment and sluggish sales, which made it tough to meet boom-era investors’ expectations.
“What’s happening is hard on a lot of good people,” says Richard Dweck, CEO of Cambridge, Mass.-based consulting firm 3rd Millennium. Still, pioneers like Dweck, who founded his company in 1994, can’t help but feel some relief now that competition for everything from talent to clients has eased.
Dweck’s business model isn’t inventive. “We’re asking the client what they want, and making sure they get it,” he says. But it’s working. “We may end up being the Bubba Gump shrimp boat of bioinformatics,” says Dweck, half-joking. He believes customization in this market is key, and that’s why consulting can be lucrative.
There is room for software vendors, too. They just have to listen to clients. “The basic thing people want is interoperability so they can do exactly what they want,” Savage says. He sits on the board of SciTegic and points to Pipeline Pilot as an example of such a product.
Other companies are “getting” that interoperability is key, including Accelrys Inc., LION bioscience, and TurboWorx Inc. (See “The Missing Links,” Bio·ITWorld, August 2002.).
Meanwhile, players such as Accelrys, SAS, and Tripos are showing that it’s easier to move down the pharmaceutical value chain than up.
Pharmacopeia-owned Accelrys started by addressing protein modeling. “It has been relatively straightforward for us to move into the sequence arena,” says Scott Kahn, general manager and senior vice president of Life Sciences at Accelrys. It’s not that bioinformatics is easy to learn. “A lot of structural biology is extremely specialized,” he says. Having a revenue base that hinges on the higher value, upstream tasks also helps, because they already have a solid revenue base, whether or not their bioinformatics initiatives catch on quickly.
Things have also gone better for groups that sell across different industries, such as data visualization companies OmniViz Inc. and Spotfire Inc., because they haven’t had to nab a lot of the pharm/biotech pie immediately to survive.
New companies are starting up -- they are just doing it the old-fashioned way. “We’re staying lean and mean,” says Michael Griffin, president and chief technology officer of Xpogen Inc. in Cambridge, Mass.. “We’ve carved out a niche in next-generation tools --going beyond things like Self Organizing Maps and K-means.” Xpogen scrapped its plans to develop a platform, but things are still going well. “In a dismal market, we have gotten a very positive response, at least from biologists,” Griffin says.
Also based in Cambridge, Biosift Inc., which formed in 2001, is targeting genomic visualization but holding on to its dream of developing a platform. “We’re starting out by showing what we can do,” says Vivian Sinno, Biosift’s director of marketing. “We knew it would be tough, but we saw a great opportunity for us, and the response is encouraging,” she says.
The most important new trend is the steady improvement in microarray data handling. Proteomics, metabolomics, and all the other “omics” that follow will benefit from this progress, accelerating the field overall.
Genomics is also being integrated into mainstream activities, and as this trend progresses, front-runners are starting to look beyond basic data integration to knowledge management, a fuzzy concept for scientists. “The main thing companies need is an integrated information environment,” says Accelrys’ Kahn. “In other fields, people call that knowledge management.”
Plenty of consolidation will happen in the months ahead, particularly in the crowded field of microarray analysis. Key things to watch will be how LION bioscience’s DiscoveryCenter launch goes and where stalwart database-provider GeneLogic stands a year from now.