RADNOR, Pa.--While bioinformatics is still a nascent science, there's plenty of room for innovators like Michael Liebman to define what, exactly, it is. Liebman, who holds a PhD in physical chemistry, teaches a bioinformatics course at Northwestern, and is writing a book on the subject, spoke with BioInform last month about his new job blazing bioinformatics trails at Wyeth-Ayerst Research here.
Last October, 50-year-old Liebman quit his job directing bioinformatics and genomics for Amoco subsidiary Vysis in Chicago, where he had worked for 10 years, to move back to his home state and create a bioinformatics department for American Home Products drug manufacturer Wyeth.
In a stark, windowless office decorated with framed photos of his wife and young daughter--who he sees on weekend commutes to Chicago--Liebman spread a series of overhead transparencies across his cluttered desk and described, as he had to several Wyeth management groups already, the strategy he spent the last three months devising. His plan puts bioinformatics at the core of Wyeth's drug discovery.
In 1996 Wyeth set out to develop a genomics presence by allying with Millenium Pharmaceuticals to gain access to that firm's bioinformatics software and proprietary "transcription filing" technology. From the small staff Wyeth installed to service the alliance, Liebman will build a department that goes beyond providing computational tools for mining genetic data and integrating clinical observations. His diverse staff will help researchers determine just what they should be looking for, he said.
"The hardest thing in science is to ask the right question, and ask it the right way" Liebman said. "A lot of users ask very specific questions without telling you what they need the answer for." Getting bioinformaticists' input early on could save researchers time, Liebman said.
"It's interesting the number of problems people thought were sequence-based that turned out to be structure-based," he said. "When we work with people we have them come in and present the background."
Composed, so far, of a computer scientist, molecular biologist, database engineer, chemical engineer, biomedical engineer, structural biologist, and a computer software engineer, with help from an oncologist and a hematologist, the bioinformatics staff will be able to look at research problems, not just sequences, and put their heads together to achieve novel solutions, said Liebman.
In fact, he said, he hopes to free his staff to do more problem solving and less routine service-oriented work by enabling drug discovery researchers to do bioinformatics on their own. The plan is to make tools and training available online. "Researchers will be doing bioinformatics instead of having someone do it for them," said Liebman. Then, bioinformaticists can branch out.
Liebman's philosophy of bioinformatics is "broader than just a genomics perspective." Until now, the science has been focused on target development, he said. But bioinformatics can help stratify diseases and model disease processes.
The challenge is "how to take bioinformatics from the viewpoint of genomics, enhance that perspective, and add a physiology view to bring [researchers] closer to the condition they're trying to treat," he said.
Inverting Drug Discovery
Bioinformatics could instigate a reversal in drug discovery methods, said Liebman. "Now, the whole process starts with target identification. We're inverting the view and saying 'this is the disease, now how do you develop validated targets?'" he explained.
This approach is the reason Liebman chooses to work with an object-oriented database instead of a relational one: "We don't know which relationships are critical. Classification limits your perspective," he said.
Mainly, Liebman's group will support discovery needs in therapeutic areas and train researchers to do bioinformatics, according to the strategic plan.
Via Wyeth's intranet, Liebman also plans to set up a system for sharing research and bioinformatics problems: "The same questions come up again and again," he said.
Liebman is developing liaisons within other Wyeth departments, collaborating with outside scientists, and actively recruiting bioinformatics staff. But bioinformatics on the scale of his nearby neighbor SmithKline Beecham is not his goal: "We're not just trying to catch up or compete with what other groups are doing. I'm positioning us to capitalize on our strengths and be somewhat unique," he said.
--Adrienne J. Burke