Part 3: CSIRO

The Commonwealth Scientific Industrial and Research Organisation (CSIRO) was set up in 1916 as the Advisory Council of Science and Industry. Over the past 80 years, CSIRO has developed a reputation for doing good, applied science that is useful to Australia. CSIRO has a number of Divisions that focus on areas of science from animal health to atmospheric research and from textile and fibre technology to tropical agriculture. All of the research groups within CSIRO are multi-disciplinary to allow the greatest exchange of ideas and assist in the commercialisation of the research being performed.

CSIRO Health Sciences and Nutrition is a division focussing on pharmaceutical and food technologies with laboratories in Adelaide and Melbourne. One of its projects recently created Caduceus: the world's largest computer dedicated to drug discovery (in fact, Caduceus is 64 computers linked together in what is called a Beowulf cluster, allowing multiple parts of the same problem to be tackled at once). Caduceus will be used to model the interactions of proteins to help find molecules that are suitable for use in new drugs and can examine up to 700,000 compounds per day.

Mr Kim Branson is a PhD student at CSIRO Health Sciences & Nutrition, and the Walter & Eliza Hall Institute (University of Melbourne). His PhD project has been funded by the Cooperative Research Centre for Cellular Growth Factors although he originally started his work at the University of Adelaide. Kim's research is aiming design (rather than discover) drugs for a specific application, but in following this aim he ended up designing and building Caduceus.

Kim's training includes a Bachelor of Science from the University of Adelaide, with an honours project in Obstetrics, Gynaecology, Microbiology and Immunology. But his life is not all science. Not unsurprisingly, Kim's hobbies include computing and creating electronic music (his current project is called "Point of Failure"), but he also has a Bachelor of Visual Arts and enjoys sculpture.

What sets CSIRO apart from other research organisations in Australia?

The finished Caduceus

In a single word, diversity. CSIRO is unlike any research institute, which may focus on a particular problem or area, such as the biology of cancer or applications of pure mathematics to telecommunications. CSIRO covers almost all aspects of science. This gives it some unique advantages. For example, if you come across a problem or result in an area which is outside your expertise and you don't feel like teaching yourself all there is to know about it, you can quite quickly contact an experienced person in that area and discuss the findings or the issue. Maybe then you might decide you wish to know more about that area. CSIRO also has a long history linked to Australian industry. I'm sure these links will continue in the future. Certainly in the past, CSIRO concentrated on things such as pest control and applied science type areas, but today you find more IT and other types of areas reflecting the problems in which today's society is interested.

Why is CSIRO a good place for a scientist to work?

Other than the multidisciplinary environment I've just talked about, the general level of skill around the place. When you are surrounded by very good people, you tend to strive to do the best you can; it helps you explore an idea properly. The feeling is one of a professional team, well-funded and with a purpose. It's nice to work with good tools - it leaves you time to focus on what you should be doing (science) rather than fixing something, or waiting to use apparatus.

Does having a focus on commercialisation impact on the sort of research you do?

Not at all, I'm interested in drug design, and applying computational methods to the study of protein/protein and small molecule interactions. I certainly don't go and tell everybody about the latest structure I might be working on, but then anyone working in an interesting area has competition on an international scale. It's a part of the game and I think there is no reason why scientists cannot make money. Why not work on something that has the potential to help a lot of people and also generate income? Unfortunately someone somewhere has to pay for the research any organisation does. But I still have some independence. I don't have any orders from anyone really to do this, or arrive at a certain time... in fact I'm quite well known for only appearing about 11:00 each day (I prefer to work till late).... The only real impact it has is knowing that a huge multinational with lots of good people and more money than you can think of might also be looking at the same problem and you have to work smarter than them not harder.

The frame on which Caduceus is built.

How are research goals determined?

The partially built Caduceus.

These things are never fixed; it's a constantly moving concept. Sure there are long term research goals one might have (a compound for receptor X), but the intermediates along the way you really set for yourself. This is based on the results you have, what's suddenly been published and what areas you are interested in. I never really had the misfortune to work under any dictatorial supervisors.

What is the role of industry in the work of CSIRO researchers?

CSIRO works closely with Australian and international industry and other government research organisations. The state of research funding is changing all across Australia - universities have increased their industry affiliations and will need to strengthen these ties if they are to survive. The same holds true for the CSIRO - the government funding is not an endless pot of gold, and realistically the CSIRO needs revenue streams from contract research for other organisations, commercialisation of its spin off companies and other successful projects. This does not mean CSIRO only does applied work for industry, rather CSIRO does the sort of basic research from which novel technologies spring. This may then be licensed to existing industries or used to found new industries. This has the potential to create new sectors in the Australian economy.

Your training is in biochemistry and immunology, yet your work on Caduceus is one of a computer scientist, how did such a transformation come about?

When I was choosing what area I was going to study in my PhD, I decided biology had finally come of age in terms of theory. We now know enough that we can do "theoretical biology" and get good agreement with experiment. I decided I'd like to work in this area, and saw it as a new field. One of the main pieces of equipment I use, as a computational biochemist is a supercomputer, so naturally you begin to learn about such beasts. Due to a long-standing interest I've had with computers this was possibly less of challenge to me. The other thing about supercomputing is that you always need more grunt - no matter how fast the computer you always wish you had something faster. As a result, you attempt different questions and want answers faster, rather than waiting 3 months for a calculation. In the end, you have two options. You could have a limited time-slice on a really expensive machine that is shared by many people and may not be really well suited to the type of work you do due to its general nature. Or you could build your own supercomputer, tailored to your particular task, and often outperforming the other more expensive option. The bonus is it's your machine so you get to use it all the time. So this is really why I designed and built Caduceus.

Inside Caduceus... what a mess!

What are the challenges of working in the CSIRO environment?

Possibly the bureaucracy, but realistically its not that bad. The fact is that some other researchers from other sectors tell us that we get it too easy at CSIRO, with so much research support. I think you need to be able to do the little things like make up your solutions etc, but in the end you are there to do the science. No one likes to do that sort of run-of-the-mill thing, it's probably a bit of jealousy really. I'm looking forward to becoming an old grouch so I can say "dammit, in my day we had to build our own supercomputers...".

« Part 2: Key Centres and Special Research Centres | Part 4: (not available yet) »