When Eric Jakobsson took a new job with the National Institutes of Health, part of the agreement was that he would split his weeks between Bethesda, Maryland, and his job at the University of Illinois, jetting back and forth between the two.

"My joke is that I'm Agent 00CB," Jakobsson said. "I have a license to kill myself in the name of computational biology."

Time will tell how the pace might wear on him, but Jakobsson has taken on the new assignment with humour and a strong sense of purpose as the first director of the Centre for Bioinformatics and Computational Biology (CBCB) at the National Institute of General Medical Sciences. The center is the newest of 27 components that make up NIH, the federal biomedical research agency. Key to his duties as director will be to help set and push a national biomedical computing agenda.

"It looks to me as though this is really the position which would have absolutely the most leverage on how biomedical computing goes in the country," he said, adding that although the initiative doesn't have much in the way of resources at the moment "leverage is all a matter of where you stand," and he wants to have a say in where funding will go.

The centre supports research and training that combine biology with computer science, mathematics, physics and engineering through areas like computer modelling, database development and analytical tools. All of that dovetails nicely with Jakobsson's areas of expertise. His research specialisation is molecular biology and mechanisms and functional significance of ion permeation in biological membranes. Jakobsson plans to continue that work at the University of Illinois, where he is a professor in the Department of Molecular and Integrative Physiology.

He has had to give up classroom teaching to take the CBCB job, but will work with university students in the laboratory in Illinois.

"I've never felt I had enough time anyway," for research, he said when asked if he thought he would be able to spend enough time on his own work now that he was dividing his weeks between two states. "This is the one thing I knew when I was talking about starting [at the centre] and the one thing I still know is that I'm not psychologically ready to give [research] up, so I hope I can keep meaningful research going."

Jakobsson also will forego academic politics at the university, but "that's not bad at all," he said. Of course, he's embarked on a whole new level of politics at the centre and "the stakes are higher" as he shifts focus to federal biomedical computing policy and initiatives. To set that stage, NIH is hosting the first Biomedical Information Science and Technology Initiative Symposium in November, which is meant to result in recommendations to NIH about future biomedical computing research, setting the national direction.

"In broadest terms, I think what we really need is a nationally distributed software engineering program. We already have lots of places where computing is contributing enormously to basic biomedical research and healthcare," he said, ticking off management of clinical trials, molecular modelling in drug design and basic biological discovery such as gene sequencing and modelling metabolic pathways as examples.

Using a computing analogy, he said that desktop PCs offered users functions that work together -- word processing, email, calendar and spread sheets -- where the "output from one is automatically input for the other... but in biomedical computing, it's not that way at all. Of course, the functionalities we're dealing with are very complex, but there will be tremendous value when, for example, we can connect the software that does tissue imaging to the software that underlies the molecular bases for whatever condition we're looking at."

One objective is to create systems where text-based documents connect seamlessly. Another is to bring biomedicine to a stage where computing power makes it easy for researchers, teachers, students and medical professionals to retrieve and analyse information, to create modelling simulation tools that help them understand the issues they are working on and researching, as well as applying their knowledge in labs and clinical settings.

Getting to that point will help achieve the objective of tearing down walls that have been built up around various scientific disciplines.

"What's really becoming more and more clear to people is that information technology cuts across boundaries that we have put up," Jakobsson said. "The boundaries we have put up in different areas of academic disciplines, scientific disciplines really have to do with a defined body of knowledge, but as soon as the knowledge of each discipline is freely available and quickly available and efficiently available to people who would like to work across those disciplines and apply knowledge, as soon as information technology exists to permit that, the boundaries really do start to crumble."

Besides, he added, "cells and organisms don't know about our disciplines."

Bioinformatics is also key in fully mining knowledge being gleaned from the human genome project, which needs powerful computational systems. "From the point of view of bioinformatics, the bottom-line statement that the human genome project makes is that it really is possible to use informatics and computation to understand how human beings function, and that you can't do it without computation and the more powerful your information technology systems are in terms of letting you retrieve data, letting you see patterns in data, letting you make models based on data, the more powerful those systems are, the more you can learn," he said.

The promise of molecular medicine won't be realised without that kind of IT power, Jakobsson said.

"I see so much excitement in this field that I think, man, I'd just really love to be around in a hundred years and see how this plays out," he said.

Meanwhile, he'll keep flying between Illinois and Maryland, and in the more immediate future will deal with the "enormous challenge" of bringing together scientists and public-policy makers from NIH and various other federal agencies. That could well prove to be the biggest task Jakobsson faces.

"Talk to me in a year or so and I'll let you know if it was an advantage to come in without the background and none of the [political] baggage or a disadvantage being so naive," he said with a laugh.