Imagine a job fair with 50 high-tech companies competing to recruit one of the handful of properly qualified scientists who bothered to show up. Sounds like a pie-in-the-sky dream, doesn't it? But according to Victor Markovitz, vice president of bioinformatics systems at Gene Logic Inc., this actually happened at a recent biotech fair. And it is more or less typical of the prevailing global job market in bioinformatics and computational biology, where there are many more headhunters than heads.
In this overview article, we define the current shortfall, explore the reasons for it, and look to the future. What, we asked academic and industrial scientists in Europe and North America, will today's students need to do to call themselves computational biologists or bioinformaticians when they graduate?
The Here and the Now
"It's a tight job market," says Inpharmatica Scientific Director Mark Swindells, with classic British understatement. Companies are hiring. Universities are hiring. And governmental agencies are increasing funding opportunities for bioinformatics research and training. Everyone is struggling to find people with the bioinformatics skills they need. But the pool of scientists holding formal qualifications in both computer science and biology is diminutive. The solution that most organizations have come up with is to recruit "bright" individuals with good qualifications in computers OR research science and a demonstrated aptitude for the other discipline. As a result, the de facto norm at the moment is for on-the-job, cross-disciplinary training.
Chris Somerville, chair of the department of plant biology at the Carnegie Institution of Washington, Stanford University [and principle investigator of the Arabidopsis information resource (TAIR)], says that industry's need for bioinformaticians is mirrored in academia, where on-the-job training is par for the course. Even so, Somerville says, there is a critical and immediate need for database experts. Not so much the computer science types who write code, he says--around Stanford, these are fairly easy (if a little expensive ...) to find. Somerville is looking for biologists who can design and build databases that usefully represent biological information, but there are few of them to be had.
The Why's and Wherefores
How did we end up with so much data and so few people trained to analyze it? Swindells points back to bioinformatics' own big bang--the explosion in data generation that began with expressed sequence tags (ESTs) several years ago. Although granting bodies provided support for the sequencing work itself, Swindells believes they were slow to follow this up with grants designed to train individuals who could operate effectively in the resulting world of computational biology. Rob Last, director of discovery genomics at Cereon Genomics, concurs: "It's not only that the dot-coms are hiring. ... It is simply that there are not enough programs spitting people out."
Although the bioinformatics funding situation is changing, there are still relatively few institutions in the U.S. or Europe that offer directed postgraduate degrees in bioinformatics, says David Wang, director of applied genomics at Motorola's Biochip Division. (Next Wave spoke with three such institutions-- The European Bioinformatics Institute, Johns Hopkins University and the University of Pennsylvania.) And given that these programs will take years to start graduating Ph.D. computational biologists, Swindells is concerned that bioinformatics companies will have to deal with the resulting 4-year "knowledge gap" on top of the tightness of the current job market.
The volume of data in public data repositories is doubling every 15 months or so, and companies are establishing their own databases of genomic and proteomic information. So, what does the bioinformatics future hold job-wise? And what can you do now to be a part of it?
Gene Logic's Markovitz believes that the current hiring trends will continue for several years. "We're in the early stages," he says. Most agree with Markovitz: The current shortage of computational biologists is going to continue for a good number of years yet. It will be 2 or 3 more years before the nascent bioinformatics programs in the States start churning out Ph.D.s, and until they do, there will still be room for people who have acquired their skills less formally. "It's fairly easy for a grad student" to do this, says the Carnegie Institution's Chris Somerville. "Just sign up for Perl and other scripting languages" to improve your chances of getting hired. Somerville also recommended that students take courses in database design and management, advice echoed by Martin Leach in his contribution to this feature.
The situation for current postdocs is less rosy, Somerville thinks. There's much less time to beef up your qualifications by taking classes, and unless you're in a position where acquiring such skills makes sense for the project you're working on, "you pretty much have to be there already," he says.
Over the longer term, as graduates of existing educational programs enter the job market, Swindells sees a two-tier hierarchy emerging. Master's programs will tend to train the individuals who gather, organize, curate, and protect the data of the in silico (i.e., computer-based) laboratory and who design and manage the tools necessary to ensure that the ever-increasing volume of data are available and can be studied and interpreted. But, says Cereon's Last, "no amount of [sequence database] searching is going to give someone the tools they need to derive biological meaning from the data." It will presumably be the job of bioinformatics Ph.D. and postdoctoral programs to train the people who can identify meaning in the data noise, derive scientific hypotheses from those data, and design the experiments and statistical analyses required to test those hypotheses.
So if you're interested in embracing the new biology, you can! Seize the moment, get yourself the qualifications and skills you need, and head out into the brave new bioinformatics world!