BALTIMORE--When the US Department of Energy terminated funding for the human genome database project at the Johns Hopkins School of Medicine here last month, Stan Letovsky, the database's director, was forced to confront a difficult but increasingly common choice for bioinformatics professionals: whether to stay in the public sector or join the growing exodus of colleagues who have traded government and university labs for private industry.
"I could probably do what I do for more money in the private sector, but that is not the only issue," Letovsky told BioInform last week, as he mulled his options. Another key consideration, he and other informatics researchers said, is the difference in research cultures found in the academic and corporate worlds, particularly the difference in the way science is practiced in university informatics labs versus in research centers in large pharmaceutical companies, who are doing much of the hiring these days. Researchers who excel in one setting, they said, may not feel comfortable--or be productive--in the other.
According to Bob Phair, a former Johns Hopkins academic who now runs his own Rockville, Md.-based consulting firm, BioInfor matics Services, one of the fundamental differences between academic-style research centers and industry labs is the degree of control researchers have over their work. "If you want to choose your research area, then there is no choice: you cannot work for industry," Phair claimed. "But if you are happy simply to work in the area of science you know, then you can make much more money working for industry. But you probably won't get to stay in that area. You should expect to change fields frequently," as companies continually refocus their efforts to strengthen the bottom line, he added.
"The real question is how much risk you are willing to take," Phair continued. "If you work for a large pharmaceutical company, you will have to do what you are told but someone else is taking the risk and responsibility for getting your funding. If you want to run your own show, then you have to be willing to take on all the risks associated with getting grants."
While agreeing that academic researchers have more freedom than most of their corporate colleagues, David Searls, who left the University of Pennsylvania in 1996 to become vice-president and director of bioinformatics at SmithKline Beecham Pharmaceuticals in Upper Merion, Pa., noted that they also face some practical constraints to using it. "There is very little to stop an academic scientist from changing direction in midstream, but abandoning a direction because you are bored is not a decision to be taken lightly," he said. Most importantly, according to Searls, such shifts could cause a scientist to lose funding because, to win coveted long-term grants, "researchers must establish a track record and show consistent results." Despite the pressure to stay on track, "in academia you can still afford to be a bit more of a lone wolf," Searls admitted, freer to pursue research that may not have an obvious commercial application--or even the interest or support of other colleagues.
In contrast, "in industry, there is a high value placed on a person's ability to work in teams and carry a group agenda forward. There are clear ground rules. Businesses specialize in certain therapeutic areas and are not interested in a finding outside those areas, even if it is biologically or medically interesting."
The lack of quarterly-earnings pressure "clearly gives academia a more open-ended mandate to ask all kinds of questions," Letovksy agreed. "A genomics company hoping to sell targets to a pharmaceutical company wants to figure out how to find the targets as quickly as possible, not spend a lot of time mining a dataset for interesting biological questions. But the public sector, he continued, "is also increasingly hamstrung in its ability to ask fundamental questions because of economic pressures."
Despite such limits, however, for years informatics professionals clustered at academic research institutions and federally funded genome centers because they were the only game in town, Letovsky and others told BioInform. Specifi cally, they were the only source of the large sets of genomic data that drove the need for informatics innovation. Today, however, Letovsky contended that the gap between the capabilities of corporate and academic research labs has narrowed. Many private companies now have high-throughput labs, if not for sequencing then at least for developing cDNA libraries. And the gap is sure to narrow even more as the emphasis shifts to functional genomics, he argued. "The gap in data-production capability has diminished; the momentum is shifting to the private sector," he said. "The public sector is pretty much played out in their ability to build new factories. I don't see NIH saying it is going to fund" large-scale functional genomic centers.
As a result of this shift, "a lot of the really good work in bioinformatics is going to be done in industry," Searls claimed. "There will still be lots of needed innovation in academia, particularly on the theoretical side, but scientists who have shown that they have the chops in computer science, and the willingness, will have the chance to get up to their necks in real data."
Searls conceded that, despite such opportunities, "it can still take some convincing to overcome a sense that they are disappearing into this black hole, or as many of my friends in academia call it, 'going over to the dark side.' But my reaction upon arriving in industry was that it wasn't as different as I might have thought." For example, he said "there is a fair amount of emphasis on publication and continuing to develop your resume. For bench scientists to advance their careers, they need to be paying attention to the same sort of things as they do in academia: publishing and peer review."
Industrial bioinformatics also shares the same commitment to discovery found in academic labs, Searls added. The high cost of drug discovery, in particular, "places an emphasis on developing pioneer drugs based on completely novel modes of action. This degree of novelty means there is a lot of healthy pressure to do novel biology, discover new genes and new systems, not just refine observations and make slight variations."
Many informatics researchers have resisted such attractions, however, and what often distinguishes them, Letovsky observed, is an ideological commitment to "open" science unhindered by the need to keep trade secrets or protect a competitive advantage. "A lot resist the financial inducements for ideological reasons," he concluded. "They want to be the open as opposed to the closed sphere."