Professor Terry Speed, winner of the 2013 Prime Minister's Prize for Science
Fighting cancer by the numbers
Terry Speed doesn’t expect to see headlines reading “Statistician cures cancer” any time soon. But he knows that the right mathematics and statistics can help researchers understand the underlying causes of cancer and reduce the need for surgery.
A mathematician and statistician, he has written elegant theoretical papers that almost no-one reads. But he has also testified in court, helped farmers and diamond miners, and given biologists statistical tools to help them cope with the genetic revolution.
Twenty years ago biologists looked at one or two genes in isolation. Today they can track thousands of genes in a single cell, but to understand the results they need tools of the kind that Terry develops.
At 70, he is focusing on techniques to sort out the thousands of differences between normal and cancer cells, moving closer to the clinic with ideas to treat cancer more efficiently, and working with industry to create a tool to determine if your thyroid growth is benign or not.
For his contribution to making sense of genomics and related technologies, the head of Bioinformatics at the Walter and Eliza Hall Institute of Medical Research, Emeritus Professor Terry Speed has been awarded the 2013 Prime Minister’s Prize for Science.
Though his office overlooks where he went to high school, and is over the road from the university where he studied as an undergraduate, Professor Terry Speed has come a long way. He now works at the Walter and Eliza Hall Institute of Medical Research (WEHI) in inner Melbourne, just metres from University High and Melbourne University.
But it has been quite a journey, including more than 20 years at the University of California, Berkeley, where he is still an emeritus professor, as well as stints at Monash University, the University of Sheffield, the University of Western Australia (UWA) and CSIRO.
Much of what drives Terry can be traced back to an unfortunate event early in his life: his lungs became infected with bacteria leading to a nasty condition known as bronchiectasis. “I stayed at home most winters, except for visits to hospital to pump gunk out of my lungs.” Finally, when he was 16 years old, surgeons operated, taking out one of the lobes of his lungs.
The direct result was a dramatic improvement in health. He successfully took up rowing the next year, and is still an inveterate runner. The indirect result was a lifelong interest in medicine and medical research.
Terry’s last year at University High School was 1960. It was the year immunologist Frank Macfarlane Burnet from nearby WEHI won the Nobel Prize in Medicine. And Terry shared the school prizes with Suzanne Cory, who was to become the Director of WEHI and would eventually entice him there.
The next year Terry started a joint medicine-science degree, but after one term he realised that medicine wasn’t for him. “I had no interest in or aptitude for the laboratory work—looking down a microscope or dissecting rats.”
So he switched to a straight science degree majoring in maths and statistics. He never lost his interest in medicine, however, undertaking an honours project on the survival of a mutant gene. “The data and reasoning of genetics are naturally appealing to mathematicians.”
At the end of his undergraduate degree he married, and a few months later took up a tutorship at Monash University and began a part-time PhD in mathematics. From there, he moved to the University of Sheffield in the UK, and then to UWA.
While at UWA, Terry began consulting widely, often working closely with researchers from CSIRO. He acquired something of a reputation as an applied statistician looking at many problems ranging from the imprisonment rates for aborigines to water use in Perth and the size distribution of Argyle diamonds. Eventually he was appointed head of the CSIRO Division of Mathematics and Statistics.
In fact, he has tended to develop theory through working on real world problems. It has never bothered him that statistics is often regarded by other disciplines as a service. “I’m the guy who says yes. I view statistics as an enabling science. I love it. It enables me to play in other people’s backyards.” Ask him what he does now and he will tell you, “I apply statistics to the fields of genetics and molecular biology. Cancer is a particular focus.”
After four and a half years in Canberra with CSIRO, in 1987 Terry was offered a faculty position in the Department of Statistics at the University of California, Berkeley; one of the best such departments in the world. From the practical CSIRO, he entered a highly academic world, where he was one of the few with administrative, or even consulting experience. Before long, he was in charge of consulting and not long after that he became chair of the department.
Berkeley allowed him to interact with many of the shakers and movers in the brave new world of genomics. He joined several groups, such as the Program for Mathematics in Molecular Biology, and went to the laboratory meetings of researchers interested in topics such as molecular evolution and the Human Genome Project.
As a result, he was one of the first to see the new technology of microarrays, where the activity levels of thousands of genes could be assessed simultaneously. Although data from this revolutionary technology was jealously guarded during its early development, he became one of the first statisticians to have access to it. As a result, he was able to develop simple statistical techniques for analysing microarray data that are still used in laboratories around the world.
Interestingly, many of the methods for which Terry is widely known do not necessarily rank highly in his own estimation. Of the microarray work, Terry says, “It wasn’t a major intellectual feat. It just happened to be the right thing at the right time.” Some of the papers of which he is most proud have almost never been cited by others.
In fact, his innate sense of fairness and his statistical skills have come together in a particular interest in criminology. He served at one time on the board of the Australian Institute of Criminology and was involved in three high-profile trials and hearings. The first came early in his career, while he was a tutor at Monash University. He testified that the trajectory of the fatal bullet made it unlikely that the last man hanged for murder in Victoria, Ronald Ryan, actually fired it. But his evidence made little difference to the outcome.
While at Berkeley, he became an expert witness at the trial of American footballer and actor, O.J. Simpson, and also testified at a high-profile scientific misconduct hearing involving a student of US Nobel Laureate, David Baltimore. The result of his brushes with the law is a conviction that a lot of forensic activity, including DNA fingerprinting, is underpinned by dubious statistical methods.
Terry always retained a link with Australia. Then, on a visit to WEHI to see his old school friend Suzanne Cory, another opportunity emerged. At the urging of molecular biologist Simon Foote, she offered him a position. “WEHI was something new and different—but it was also a bit like Berkeley, offering a broad range of scientific interaction with really good people.” And it would bring him back right into the heart of medical research.
So in 1997, he began at WEHI in a 50:50 appointment with Berkeley. “For the next 13 years, I spent half my time in Berkeley, half my time in Melbourne, and the other half flying between the two.” He now works full-time at WEHI. “When I began it was a totally new world. I had no idea what a bioinformatician was. I didn’t even like the word. But I found there were plenty of people wanting to talk to me.”
He now heads a Bioinformatics Division that has developed techniques that are used in medical research all over the world. Even so, he is concerned at the precarious nature of bioinformatics in Australia. “There are lots of people doing things, but jobs tend to be of one or two years’ duration. And there are still no established career paths—so pathetically few can think about continuing to do research in the area.”
He describes his interests as “reading, writing, ‘rithmetic, and running”, but he also enjoys going out to the opera or a concert.
Terry Speed’s achievements are many, and traces of his influence span the world, in terms of contributions to statistical theory and practice—and particularly his students. In a preface to Selected Works of Terry Speed, published in 2011, his student Professor Sandrine Dudoit of Berkeley writes, “In addition to shedding light on Terry’s scientific achievements, the [accompanying] commentaries reveal endearing aspects of his personality, such as his intellectual curiosity, energy, humour, and generosity… Terry [is] an avid and tireless scholar... [His] thirst for knowledge has not abated, and neither has his dynamic pace.”
- 2005 - Hon DSc, University of Western Australia
- 1969 - Diploma of Education, Monash University
- 1969 - PhD (Mathematics), Monash University
- 1965 - Bachelor of Science (Honours), University of Melbourne
- 2013 - Fellow, Royal Society of London
- 2012 - Victoria Prize for Science and Innovation (Life Sciences)
- 2011 - NHMRC Program Grant: Interaction of malaria parasites with the host: disease, pathogenesis and control
- 2010 - ongoing - Emeritus Professor, Department of Statistics, University of California, Berkeley
- 2009 - NHMRC Australian Fellowship
- 2007 - NHMRC Achievement Award for Excellence in Health and Medical Research
- 2006 - ongoing - Head, Bioinformatics Division, The Walter and Eliza Hall Institute of Medical Research (WEHI)
- 2004 - 2010 - Executive Committee member and Partner Investigator, Australian Research Council Centre for Kangaroo Genomics
- 2003 - 2004 - President, Institute of Mathematical Statistics
- 2003 - Australian Government Centenary Award Medal
- 2003 - Moyal Medal, Macquarie University
- 2002 - Pitman Medal, Statistical Society of Australia
- 2001 - Fellow, Australian Academy of Science
- 1997 - 2005 - Joint Head, Genetics and Bioinformatics Division, WEHI
- 1990 - Fellow, American Association for the Advancement of Science
- 1989 - Fellow, American Statistical Association
- 1987 - 2009 - Professor, Department of Statistics, University of California, Berkeley
- 1984 - Fellow, Institute of Mathematical Statistics
- 1983 - 1987 - Chief, Division of Mathematics and Statistics, CSIRO
- 1974 - 1982 - Associate Professor/Professor, Department of Mathematics, University of Western Australia
- 1969 - 1973 - Lecturer, Department of Probability and Statistics, University of Sheffield
- 1965 - 1969 - Tutor/Senior Tutor/Lecturer, Department of Mathematics, Monash University