YouTube Facebook LinkedIn Google+ Twitter Xinginstagram rss  

On the Threshold of Systems Biology

By John Russell

Sept. 18, 2006 | Almost exactly one year ago, David de Graaf became Pfizer’s first director of systems biology — whatever that means. Even David wasn’t precisely sure. Ostensibly, it was to help Pfizer better understand system biology’s potential and to act as a conduit for the technology to researchers. He told me half-jokingly, if I’m still here in a year, that will be one measure of systems biology’s traction in Pfizer.

David is still there. In fact, since then, Pfizer has talked a little about one program that used systems biology methods to suggest a method of action for drug-induced vascular injury and may lead to definitive biomarkers. Last month, following a panel discussion, Bruce Gomes who works with David told me their biggest problem now was meeting increasing demand inside Pfizer. He said results of another program — he declined to specify details — had excited Pfizer researchers, and the spiked interest was stretching their ability to respond.

Pfizer is not alone. Last March, Novartis elevated its modeling effort to departmental status. The FDA, under Bob Powell, is slowly building a bank of models based on legacy data (talk about an underused gold mine) to help sponsors use simulation to design better Phase III trials. This year Dana-Farber opened a Center for Cancer Systems Biology. This summer, the American Diabetes Association partnered with Entelos to provide academic researchers with free access to a diabetes simulation platform. For years the Institute for Systems Biology has been pumping out experimental and informatics tools to deal with the omic data deluge. The list goes on.

Are we nearing an inflection point for systems biology? Hopefully. But abrupt change in the pharmaceutical industry is still a years-long process. What seems clear is that among early adopters like Pfizer, systems biology concepts and tools are becoming entrenched. They believe systems biology will speed efforts to understand biology, identify biomarkers, deliver drugs, and even help clinicians choose optimal therapies for patients based on the patient’s particular attributes.

In this issue, guest columnist Nat Goodman, a researcher at the Institute for Systems Biology, invites readers on a journey of several columns to explore systems biology. Among other things, he will review real tools that researchers can use, an exercise that is now badly needed.

In his opening column, Goodman suggests too many people still mistake the “systems” in systems biology to mean it’s mostly an in silico and theoretical endeavor. That’s wrong, he says. I agree. “Systems biology is squarely an experimental field that eats, drinks, and breathes data. To do systems biology you need an experimental system that is amenable to large-scale experimentation,” he argues. That’s certainly true today and will be for years. Not enough biology is known.

‘All Knowable Knowledge’
The promise of systems biology is large, and starting to be realized. I’m a believer. But I really like an observation by Ajay Royyuru during a conversation we had recently. He’s the senior manager, computational biology center, at IBM Research’s Watson lab. A molecular biologist by training, he did a postdoc in structural biology at Memorial Sloan-Kettering Cancer Center, and joined IBM in 1998 after a brief stint at Accelrys. Blue Gene, IBM’s supercomputer, is one of his toys.

“Systems biology efforts in general are trying to define all knowable knowledge of biological systems through a variety of platform technologies. So one often wonders what amount of work will it take to describe a complex biological system, to dissect and know the entirety of a complex biological system, and when do we know that it has been accomplished well,” said Royyuru.

“From complexity analysis it seems obvious that you can claim that you’ve understood it to some extent only when you are able to describe the behavior using tools and techniques that are far less cumbersome than just reproducing the system itself. If my in silico model ends up being as complex and does not provide any more detail and insight than the functioning biological system in the Petri dish, then have I just mimicked the complexity or have I understood the complexity? You know what I mean?”

I do. We still have a long and exciting way to go.

John Russell, Bio-IT World’s executive editor, writes a monthly systems biology newsletter. Subscribe now at, or you can e-mail John at

Click here to login and leave a comment.  


Add Comment

Text Only 2000 character limit

Page 1 of 1

For reprints and/or copyright permission, please contact Angela Parsons, 781.972.5467.