Retooling for Bioinformatics

Nurturing a career in bioinformatics takes a certain kind of aptitude, despite its popularity. First off, you have to like being intimately involved with computers, agree those who already work in this burgeoning field. "For some, staring at a glowing, blue screen all day is like a ring of Dante's inferno," says Larry Hunter, past president of the International Society of Computational Biology (ISCB) and director of the Center for Computational Pharmacology at the University of Colorado School o

Nov 27, 2000
Karen Young Kreeger

Nurturing a career in bioinformatics takes a certain kind of aptitude, despite its popularity. First off, you have to like being intimately involved with computers, agree those who already work in this burgeoning field. "For some, staring at a glowing, blue screen all day is like a ring of Dante's inferno," says Larry Hunter, past president of the International Society of Computational Biology (ISCB) and director of the Center for Computational Pharmacology at the University of Colorado School of Medicine. Number two, he says, it helps to like math. "A lot of bioinformatics involves statistics."

To retool for a job in bioinformatics first requires some considerations. You need to ask yourself: "What kind of job do I want? Research at a pharmaceutical firm or genomics start-up? Teaching and research in academia? Or, do I really want to learn the latest tools to enhance my molecular biology research?"

The most flexibility for jobseekers, says Hunter, is in pharmaceutical firms and biotech companies because the demand is the greatest there. At the very least when looking for a position in bioinformatics you'll need to exhibit an interest and basic skills, which can be picked up through mini-courses, summer courses, and other venues, to land an entry-level job at a larger company. It also helps if your biological expertise is in an area in which the firm is interested, such as G-coupled receptor proteins. But, he cautions, "funding issues alone are not enough of a motivation," to move into bioinformatics.

Specialists emphasize that there's a big difference between using informatics tools and inventing new ones. If you're serious about inventing new tools or about a more senior-level position, then you'll need some formal training in computer science, says Hunter. He adds that this is not just about programming. "You'll need to know about data structures, computational complexity, and numeric methods."

Ways to retool take on many forms: short courses and workshops, graduate and undergraduate courses and degree programs, online courses, postdocs, and sabbaticals. Degree programs are a good idea, say educators, if you have the time and money. There's a huge demand for people with the combination of a master's degree in computer science and a Ph.D. in a life science, says Hunter. A postdoc in bioinformatics is also a viable avenue, but these are competitive.

 

Living Example

Deborah Pinney's current situation fits all of these. She's a postdoc at the University of Pennsylvania Center for Bioinformatics and a master's student in computer science, also at Penn. Pinney received her Ph.D. in virology from the University of Virginia in 1984, did a life science postdoc, after which she was a research associate at the Fox Chase Cancer Center and at Penn, both in Philadelphia.

Why this direction? She was a math major as an undergrad and had always liked this subject, but she was disillusioned by what she was doing, especially the bench work. To investigate other avenues, Pinney took some computer courses until her gradual move toward bioinformatics coalesced in its present form.

In her experience, she found that she needed to get computer skills and database experience--"some pretty hardcore stuff," as she describes it. For retooling, she recommends that life scientists gain practical experience by looking for industry-based summer internships or ones at academic-based bioinformatics centers. For several months last year, Pinney volunteered as a database annotator 10 hours per week at the Penn center. This involved inputting information into a gene expression database, verifying entries, and adding new information.

For More Information

Canadian Bioinformatics Workshops
www.bioinformatics.ca

GeneEd
www.geneed.com/courses

International Society for Computational Biology
www.iscb.org

National Center for Biotechnology Information
www.ncbi.nlm.nih.gov

"There are ways to get practical experience, which you need to have on your resume to be credible," she says. "Bioinformatics isn't just knowing your Mac inside and out, or doing BLAST searches from the Web, but many biologists think it is. They really need more experience than that to succeed in bioinformatics."

 

Focusing Your Efforts

David Jensen, managing director of Search Masters International Inc. in Sedona, Ariz., runs into a lot of jobseekers interested in bioinformatics at the various career fairs he attends. He says that retooling is a process that takes one year at a minimum. "Two years would just be about right," he adds.

Bioinformatics can be broken down into two broad areas where careers can be focused: learning the development tools of the trade, which includes programming in such relevant languages as PERL and Java, and learning the analysis tools such as BLAST, which access various databases such as GenBank.

The answers to how best retool depend on where you find yourself in your career and how busy you are, advises Michael Zuker, a professor of mathematical sciences at Rensselaer Polytechnic Institute who manages the training information page for the ISCB Web site. "Some schools offer undergrad programs, but I wouldn't recommend that for everyone."

For those in grad school, he says, there are places, which are listed on the ISCB Web site, that have specific bioinformatics programs. "So if you're young enough, go get the degree in the first place." While there are certificate programs available now in a number of universities, there are only a handful of bioinformatics programs that have been going for more than one to two years.

On the other hand, if you're just about to finish your Ph.D., don't do another degree, do a bioinformatics postdoc and get immersed in the area, recommends Zuker. "But you'll have to work hard and maybe have to take some extra courses."

If you're in the middle part of your career, a sabbatical in industry or at an academic center or lab might be the way to go. In fact, Zuker took a sabbatical from 1991 to 1992 in the lab of Sam Karlin at Stanford University, where they worked on the analysis of DNA and protein sequence data, as well as protein structure.

Bioinformatics specialists say that all courses no matter where they're located--in cyberspace or on campus--are very popular. Workshops "are certainly good, especially if you only have a short amount of time, say a week or so," says Zuker. But he warns that some online courses are not appropriate for retooling: "It's like using a dictionary, you need to know the language already. They're not really for initial training."

GeneEd, an online learning center based in San Francisco for classes in biotechnology and genomics, including bioinformatics, just launched Bioinformatics 2000: An Overview, a one-plus-hour online course that gives a nontechnical overview of bioinformatics, in association with Scientific American. The company is also planning to unveil another class, Advanced Bioinformatics, a three-hour course for people who want more.

Image: Courtesy GeneEd

One of GeneEd's newly launched online course in Bioinformatics
This is an "ad hoc way" to get into bioinformatics, explains Sunil Maulik, a biophysicist and GeneEd's chairman and CEO who has been teaching biotech-related courses at California universities for the last 10 years. "This is a way to complement what universities offer. For example, we don't go into theories or algorithms. It's a more pragmatic, hands-on, applications-based approach for employees in companies. Universities are better for deep learning."

"The field is not necessarily to the point where people drop out of school to retool," sums up Cynthia Gibas, assistant professor of biology at Virginia Tech in Blacksburg, Va. "It's more helpful for biologists to get formal computer training than vice versa." To her, retooling in a nutshell means learning how to do BLAST properly and learning to be comfortable with computers. Gibas teaches seniors and some graduate students BLAST, gene finding within a sequence, multiple sequence alignment, sequence profiles for constructing phylogenetic trees, and protein structural prediction. Her book, Developing Bioinformatics Computer Skills (Cynthia Gibas, Per Jambeck, O'Reilly & Associates, New York, 2001) is due out early next year.

"It is indeed a very exciting area for a career," notes Jensen. "But, I'm often asked: 'Is this the solution to my problems?' by those seeking a career that will always have marketability. I don't think that anyone can make this promise. That's because right now there are many, many people currently retooling for bioinformatics. Someday the saturation point will come, as it did for molecular biology." But he adds, determining how far off that will be is anyone's guess.

  Karen Young Kreeger (kykreeger@aol.com) is a contributing editor for The Scientist.