1. With an educational background at Monash University where you obtained your PhD, when and how did you first become aware of statistics as a discipline?

I first became aware of statistics at high school. I didn’t learn statistics at that time, as I did the “hard” maths courses which had no statistics in them, but a close friend took the “easy” maths course, where they did some statistics. He showed me what they were learning. I saw the normal curve with 1/√2π in it, and got fascinated by that. I still am.

2. What was it that inspired you to pursue a career in statistics?

No single event or person. I found that I liked it and was o.k. at it. But I did spend some time trying (and failing) to be a mathematician, then probabilist, before I “found my niche.”

I started off doing medicine at the University of Melbourne but I didn’t do very well in the lab work. So, after one semester, I switched to mathematics and statistics. My original goal was to do medical research, and that is what I am doing now. The University of Melbourne is across the road from where I work now, and also next door to where I went to high school. For what it’s worth, it’s also across the road from where I live! I did go away for thirty years so it’s not as embarrassingly parochial as it sounds!

Anyway, after taking my undergraduate degree, I wanted to study for a PhD in Statistics at Berkeley but that did not work out. They offered me some money, but not enough to support my wife so I couldn’t go. That was 1964. I eventually got a job at Berkeley twenty years later. So, instead of going to Berkeley to do a PhD in Statistics, I went to Monash to study for a PhD in mathematics. I’d nearly left for a real job before starting research, but my professor persuaded me to stay in the academic world. After my PhD, I taught at University of Sheffield for four years but my wife did not like living there much and asked if we could move back to Australia. We ended up in Perth and stayed there for 9 years. I was teaching in the Department of Mathematics, where I became Professor of Statistics, and eventually Head of the department for a couple of years, before leaving for a (government) job in CSIRO, to be head of their mathematics and statistics division.

…it hasn’t been a straightforward path. I tried to be a mathematician and failed, and then I became a statistician, and later a bioinformatician.

Then after four and a half years, I left for Berkeley and ten years after that, I left to work 50% where I am now at the Hall Institute. During my last year of high school, I was the second to top student and the top student at that time was later the Director of the Institute. She asked me in 1996 if I could come and work with her, and my wife said yes!

So it hasn’t been a straightforward path. I tried to be a mathematician and failed, and then I became a statistician, and later a bioinformatician (which I think is a dreadful word), but you may well say my core competency is statistics.

There wasn’t anyone in particular who inspired me to pursue a career in statistics, but there were always people whose work I liked and learned from. I like to figure out things for myself, so I don’t really have people or events that have inspired me, but have just pursued my own goals. Australia had a very strong statistical tradition when I was a student, so there were lots of “role models”. I would read their work and attend their lectures, and I was inspired by them in that way. I’m embarrassed to say that most of them were male, but I did once co-author an article (with my wife and two others, one a woman) about three pioneering women statisticians in Australia. It was never 100% a male preserve, despite appearances to the contrary.

3. You are currently Head of Bioinformatics at the Walter and Eliza Hall Institute of Medical Research at Melbourne. Please could you tell us more about your role and your work there? Does the Institute have any plans with it being the International Year of Statistics?

When I was asked to join WEHI, I started bioinformatics, i.e. there was just me. There are now about 40 of us: four senior scientists, a number of postdocs and research officers, and we have students working with us too. Our general goal is to consult, collaborate and to publish enough research in order to continue our funding. It is a challenging life compared to academia, where you are guaranteed a salary. At the Institute, you can be doing well, but perhaps not well enough to obtain a fellowship, or the renewal of one if you get one initially.

We do a mixture of collaborative work with our scientists and those in other institutes. We have to watch what we do to maintain our funding, doing some of our own work as well as in collaborations, in order to demonstrate our capacity for independent research. There are buzz words, like first author and last author on papers; you don’t want to be a middle author all the time, you need to show you can be an initiator. Funding is a very important issue in our medical environment, but still I must apologize for mentioning it so often.

I enjoy developing methods that have not been devised yet. I don’t give advice, but if I did, I would say, ‘Jump in early.’ If you are on to something at the beginning, you will make a lot more impact, in comparison with developing work further that has already long been established.

Having said all that, it’s a great environment to work in. We see lots of interesting problems and data, and meet enthusiastic colleagues, people who are also in the same boat as us. I enjoy developing methods that have not been devised yet. I don’t give advice, but if I did, I would say, ‘Jump in early.’ If you are on to something at the beginning, you will make a lot more impact, in comparison with developing work further that has already long been established.

Less important for the Institute but important for us, is how we’re regarded by the Health and Medical Research Council.

We have no plans for the International Year of Statistics at the Institute itself, but I am giving a few lectures this year in honour of the event. At WEHI we are not perceived as statisticians although some of us are. The background to bioinformaticians can be quite varied – we have physicists, biologists, computer scientists, etc. Two of our staff have PhDs in mainstream statistics and one in mathematical and statistical genetics. We also have a staff member with a background in astrophysics. I’m a statistician and it’s probably the dominant way of thinking in our group.

4. You were President of the Institute of Mathematical Statistics in 2004. What are your memories when you look back on this time and what do you feel were your main achievements?

I tried to do a little for women in our profession. I wanted the IMS to find childcare facilities in or near the conference centres we used, and subsidise people so that they could bring their children to our meetings. Even with the best will in the world, this is complicated, but it is ticking away. In the same vein, I am working with others to put in place a childcare facility at my Institute.

One of my columns as President for the IMS Bulletin was entitled ‘Keep Gender on the Agenda’ where I started by explaining how in my mother’s generation, most women were expected to resign from their job once they were married. F.E. (Betty) Allan, an early Australian biometrician who studied with Fisher and wrote a joint paper with John Wishart, was forced to resign from her government position when she married in 1940. A lot has changed since then and now roughly half of the graduate students I have worked with at Berkeley since 1987 have been women. But a lot still needs to be done towards developing their later careers. I wrote ‘The purpose of the IMS is to foster the development and dissemination of the theory and applications of statistics and probability, any development and dissemination we do involves people’. With 15% of our members being women, I asked that we worked hard to ensure that women are more involved and to work together to make sure that 15% became 50%. I was not asking for a radical change as the issue was and still is not a shortage of qualified women – it was getting them into the field of vision of many men.

5. How do you think the Institute of Mathematical Statistics has evolved during your time there overall and adapted to the changing needs of the statistical community?

Generally very well. I suppose it may appear unusual to talk about women in statistics for a presidential address when most would talk about something more on the theoretical side and which statistical methods have contributed to various areas of research. I took quotes from the film Casablanca to illustrate various different aspects of the situation, such as the most famous “I think this is the beginning of a beautiful friendship”, advising that mentors and networking are both important to women and to men, and that we needed to cast our net as wide as possible when we sought speakers and organizers for our meetings. I had one year to make a difference, and I tried.

‘The purpose of the IMS is to foster the development and dissemination of the theory and applications of statistics and probability, any development and dissemination we do involves people’. With 15% of our members being women, I asked that we worked hard to ensure that women are more involved.

I’m not very good at looking back and considering what my achievements were as I tend to look back and find failures, probably because I always wanted to do more at the time. Let’s leave the achievements to someone else to think on when I’ve passed on!

I have many happy memories at the Society; it’s not too enormous an organization and is really responsive to their membership. One year of Presidency is not enough to change 80 years of history – you try and do what you can during that time.

6. How did you come to be an expert witness in the O.J. Simpson trial? Did your testimony involve statistics?

Yes to the last bit. I was then and still am concerned about the use of statistical evidence in DNA forensics, and at the time this was well known, so I was a natural choice to ask to serve in a pre-trial hearing on the use of statistics. I’ll explain how that morphed into my appearing for the defence.

I’m sceptical of the way in which statistics is used in DNA forensics. I expressed my views at some conferences and meetings and became known as a statistician critical of the prevailing consensus. The statements that can be made in the world of DNA concerning the strength of evidence use phrases with incredible numbers such as “100 million to one chance”. This is not scientifically founded and gives a thoroughly misleading view on the strength of the evidence. There continues to be debate about it to this day.

I’ve argued in conferences and in testimonies in court that forensics does have logic to it but it is not only science nor is it only statistics. In Texas a few years ago, it became evident that a laboratory assistant did not analyse the data and made up the answers and as a result, all her cases were retried.

At the beginning of the Simpson trial, there was going to be a pre-trial hearing and experts from both sides would argue in front of the judge as to what approaches should be accepted. Other pre-trial activities dragged on, and the one on DNA forensics was eventually scrapped. The DNA experts, including me were then asked whether they wanted to give evidence for the prosecution or defence, or leave. I did not initially plan to join the defence team, but wished to express my point of view in what was more or less a scientific environment before the trial started, but when the pre-trial DNA hearing was scrapped, I decided that I had no choice but to express my views in court on behalf of the defence, which I did.

The statements that can be made in the world of DNA concerning the strength of evidence use phrases with incredible numbers such as “100 million to one chance”. This is not scientifically founded and gives a thoroughly misleading view on the strength of the evidence. There continues to be debate about it to this day.

I was not a very important witness and there was a very astute lawyer on the defence team who found a mistake in the statistical calculations of the prosecution, one which I had not found. I was there to try and point out some of the defects in the way things were done, mainly relating to the strength of the evidence and proficiency testing.

7. You have developed methods of analysis now in daily use in research laboratories worldwide e.g. scientists can now accurately tell which genes are being turned on in a cell, how much each gene is being turned on and what sort of transcripts are being produced. Is there a particular area of research you are most proud?

I have a few papers explaining very simple methods that seem to work, and get widely used and cited. That is one kind of satisfaction; call it “worldly acclaim.” I got used to that idea later in life. About 15 years ago I was asked to list my “best” ten papers. I did, and it was pointed out to me that several had no (i.e. zero = 0) citations! Clearly, they were the ones that gave me personally most satisfaction, usually because I found it hard to do that research, and was pleased when I succeeded. But that doesn’t always or even usually mean the rest of the world cares or notices.

Nowadays, I tend to like things that are transparent and understandable, and as I’ve got older, I’ve liked simpler methods that do the job. A lot of what people do in statistics is driven by theory. I’m not against theory at all, but I now prefer a more practical perspective. I think you need an external assessment to tell whether a method does the job, independent of statistics (theory, simulation), and if the external assessment shows that a method does the job, that this is accepted by users who are biologists and scientists, and not just statisticians, then you’ve made your point.

8. You are best known for your contributions to the analysis of variance and bioinformatics and in particular to the analysis of microarrays data. What is it about these areas that fascinate you?

I am fascinated by engineering and gadgets and microarrays provide data on the activity of cells. It’s a complicated process to consider and I am also fascinated by measurement. Biological measurements have lots of interesting features. When you’re a statistician, you need to understand the data, e.g. with survey research, how it was put together in the first place, reliability of responses, and so on. Microarray statistics is the analysis of data put together by lasers firing at tiny pieces of DNA stuck on a microchip. As you can imagine, this a very complex process and there is lots to tease out, and I find the whole thing fascinating.

Modern labs are packed with machines that generate data. To some statisticians, a number is a number, but to me, a number is packed with history. If someone takes a blood sample from you, and later they measure the cholesterol and haemoglobin count, I get interested how each of those different measurements is obtained – what is the impact of the results in terms of were you sitting or standing when you gave the blood? What temperature was the blood at the time? You need to understand a fair amount of history before beginning to interpret the statistics. I was giving a lecture recently to biologists and I was explaining about giving blood, where they test your blood count first. I have been a regular blood donor for decades until one day I tried and was informed that I was anaemic, i.e. my hemoglobin was too low. I explained I should be fine and asked if it could be done again, and they did so, and the next result was over the required level by quite a bit! The nurse tried again a few times with the same sample a few more times and the results were all over the place! This is an example illustrating my interest in the process of measurement, and what it can come up with. There are lots of stories like this: it’s always important to understand the measurement process.

9. Your research has focussed on Metabolic flux analysis; Estimating 13C enrichment in time course experiments; Base calling for re-sequencing chips and Phylogenomics Illuminates the Evolution of the Apicomplexan Phylum. What are your main objectives and what do you hope to achieve through your current research?

The phylum Apicomplexa contains the parasites that cause malaria and we have a group here in Melbourne who studies malaria. I have worked closely with them over the years. I am interested in the genomes and evolutionary history of the Plasmodium species, which are the parasites which causes malaria.

For decades, people have been working on a vaccine to prevent malaria and this is still an interest.

To some statisticians, a number is a number, but to me, a number is packed with history. If someone takes a blood sample from you, and later they measure the cholesterol and haemoglobin count, I get interested how each of those different measurements is obtained – what is the impact of the results in terms of were you sitting or standing when you gave the blood?

10. Do you have any advice for students considering a university degree in statistics?

I try to avoid giving advice directly. I’m always happy to chat to individual students about their interests, hopes and plans, and in the course of that, make comments and/or judgements and reflect on my views and experience. I much prefer this to “advice”.

11. Over the years, how has your teaching, consulting, and research motivated and influenced each other? Do you continue to get research ideas from bioinformatics and incorporate your ideas into your teaching? Where do you get inspiration for your research projects and books?

All of the above. Not easy to speak about without getting specific. If you are teaching, you should be practicing. You cannot be divorced from the practice of research.

I get my research ideas by being asked to do something that hasn’t been done already. These days I tend to apply myself to more practical ideas. Once I’m onto an idea, it’s like a disease in a sense: I’ve caught it and have to work through it!

I am happy to talk to anyone, and even boring people or things can turn out to be interesting sometimes!

12. What has been the most exciting development that you have worked on in bioinformatics during your career?

I tend to get excited with and absorbed in whatever I’m doing at any given time. I think the thing you are getting at is all about looking back, and then it’s not excitement, but small satisfaction/ big weariness.

13. What do you see as the greatest challenges facing the profession of bioinformatics in the coming years?

Big Data is of course an issue and managing the data is an issue. I find it hard to answer a question like this.

14. Are there people or events that have been influential in your career?

Plenty of people: great ones (dead or alive), contemporaries, competitors. Not so many events, at least not as things are happening. Pearson, Fisher, Neyman, Tukey, all these and other greats, and I include Alan James from Adelaide, and many friends and competitors: all of these influenced me when I was younger.