Harvard’s three-year-old systems biology department seeks to define a new way of doing a new science. But is it the best way?

Robin Orwant

Angela Reese’s job as a staff assistant in Harvard Medical School’s new Department of Systems Biology includes an unusual task. Moments before a departmental seminar begins, she winds her way through the fourth and fifth floors of the school’s Warren Alpert building, pausing at each lab in the department to bang a gong.

“It’s a way to get people’s attention and bring them all together,” she explains.

This unusual approach of bringing people together is emblematic of how unusual the department is and what it’s trying to do. Less than three years old, the systems biology department is Harvard’s first new basic science department in 20 years. Not everyone is convinced it deserves to be there, but with more than two dozen faculty and a graduate program in place, it’s clearly here to stay. Researchers from a variety of backgrounds are coming to the department to try to define the field and train what they hope will be a new breed of interdisciplinary scientist. And they face many hurdles. With so much buzz about systems biology, other institutions are closely watching Harvard’s high-profile example. But it’s debatable whether Harvard’s department is the best example for others to follow.

A major challenge for the young department—and for the fledgling field—is to define systems biology. “No one knows what systems biology is, of course,” joked Stuart Kauffman, a systems biologist at the University of Calgary, during a recent talk he gave to the department. A few chuckles rose from the crowd of 80 or so crammed into a conference room.

What most can agree on is that systems biologists try to think about a biological system as a whole rather than focusing on just its constituent parts. With the help of high-throughput technologies like gene chips, they seek to understand how all these parts interact and fit together.

“Instead of thinking about one gene, you can think about all genes or all proteins,” says Pam Silver, one of the department’s founding faculty.

But how do you define a system? Is it a whole organism? A cell? An organelle? A biochemical pathway? Do two interacting proteins count as a system? Depending on how narrowly you define it, almost any biologist could say he or she is doing “systems” biology.

Another problem introduced by systems biology is how to deal with such large amounts of information. “If you just have your protein and you want to know, Does it bind DNA?, you can wrap your mind around it,” says Silver. “But if you’re looking at many things, like all genes in the genome, that inevitably involves a lot more computation.”

This has led to the growing use of computers—a defining feature of systems biology. Computers become even more crucial as researchers incorporate these data sets into mathematical models that try to explain how a system works.

Even more daunting is the breadth of knowledge—in mathematics, biology, physics, computer science, and other fields—that seems to be required to pull this off. Collaborations between such different fields are notoriously difficult to form and maintain.

Quantitative scientists have a very different mentality compared to biologists, says Alex Mogilner, a mathematician at the University of California, Davis, who works with biologists. “Mathematicians want elegance,” he says. “Biology is messy.” These differences can lead would-be collaborators in opposite directions unless they force themselves to depart from their usual ways of thinking.

The Harvard way

In a sense, Harvard’s new department exists to address these sorts of problems. Harvard is trying to tell the world what it thinks systems biology is and how it ought to be done.

And people are paying attention. Kauffman, who is a leader in the field and is spearheading a systems biology initiative in Canada, told department members, “It’s wonderful what Harvard is doing! We hold you guys up as the exemplar.”

While there were already several systems biology institutes, centers, and research groups throughout the world when Harvard formed its department, Harvard stands out for its long-term commitment to the field.

“It becomes a much more serious undertaking if you’re setting up a department,” says founding chair, Marc Kirschner. “This is not something that the present dean can start and a future dean can stop.”

That seems to be a selling point. Of his decision to join the new department, Vamsi Mootha, who won a 2004 MacArthur “genius” grant for his work on diabetes, says, “the fact that there was this level of commitment to form a department of systems biology was very attractive.”

Apparently, a lot of other people agree. The faculty has exploded in number from the original four to 29. A recent high-profile recruit is Peter Sorger, the former director of the Computational and Systems Biology Initiative at MIT.

The only way?

The big question, though, is whether Harvard’s department can serve as a model for how systems biology research should be done elsewhere. Most people agree that systems biology requires interdisciplinary collaboration. But not everyone is convinced such collaboration will happen in the new department. All the elements are there: chemists, physicists, electrical engineers, mathematicians, theoreticians, and computer scientists appear ready to work together. Appearances, however, can be deceiving.

“If you want to collaborate, you have to step down from your ego,” says Gaudenz Danuser, a professor of computational cell biology at Scripps Research Institute in La Jolla, CA. He’s an engineer by training, working in the institute’s cell biology department. “I’m very curious to see how these superstars [at Harvard] collaborate with each other.”

There are also geographic barriers to collaboration. Although the core of the department is based at Harvard Medical School, the faculty members are spread out across Boston and Cambridge. Mootha’s lab, for example, is at Massachusetts General Hospital. Others are based at hospitals in the Longwood area or across the river at the main Harvard University campus. Many have joint appointments with other departments or other institutions. Though they do get together for faculty meetings and departmental retreats, it isn’t clear how interactive the “off-quad” faculty members are with those on the medical campus.

“The reality of the commute across the river is not trivial,” Kirschner admits.

But even if the university unites the entire department on its new Allston campus, as some have hinted, it may cause as many problems as it solves. “By putting us together in a place like Allston, we lose the connections over here [at the medical school] and they [systems biologists not based at the medical school] lose the connections they have with the kinds of departments they’re in,” says Kirschner. “Wherever we end up, we’re going to always be confronted with the fact that we have roots in so many different entities and we’re going to have to maintain those roots.”

Teaching systems biology

Perhaps the most intriguing feature of Harvard’s systems biology department is its graduate program. Eight graduate students are completing their first year and another eight will join their ranks in the fall. Many of them have experience both in biology and in a more quantitative field at the undergraduate level—a definite help.

It’s not clear whether Harvard’s method of educating the next generation of systems biologists is applicable at other universities. Harvard gives its systems biology students a great deal of freedom in choosing which courses to take. Students can tailor their coursework to meet their perceived weaknesses.

This kind of flexible program probably won’t work elsewhere, says Mogilner. Typical graduate students may need more direction. “When it goes mainstream, there should be a very well-defined curriculum,” he says. But what should that curriculum be? “Nobody really knows what skills a systems biologist needs to have,” says Mogilner.

Whether these Harvard students will become the leaders of a systems biology revolution is impossible to tell right now. “I think we’ll only know when the product emerges,” says Kirschner. “As I’ve said from the beginning, I cannot predict that this is going to bring results in the next five years, in the next 20 years. I have no idea.” But he hopes the department is planting the seeds that could, if all goes well, lead to a new way of thinking about and doing biology. And that’s something that will only be clear in retrospect.

Robin Orwant is a freelance writer in Brookline, MA.