Thanks for visiting Scientific Computing World.

You're trying to access an editorial feature that is only available to logged in, registered users of Scientific Computing World. Registering is completely free, so why not sign up with us?

By registering, as well as being able to browse all content on the site without further interruption, you'll also have the option to receive our magazine (multiple times a year) and our email newsletters.

Computing, cricket and genetic variation

Share this on social media:

Topic tags: 

John Murphy profiles a leading statistician of our DNA

Many people do biology because they can't handle maths. This has become a bit of a problem, because so much of modern biology is about maths. The science of genomics leads to vast quantities of very accurate data being produced, and there is a pressing need for sophisticated mathematical techniques to make sense of it all.

Peter Donnelly, professor of statistics at Oxford University, is one of the foremost researchers in that very field. His background is in mathematics and statistics but he has migrated to working on how this data can be handled so that science can link what comes out of genome research to real problems in the world, such as alleviating or treating human disease. A very few diseases are known to be caused by a specific gene, but most have less direct connections to genes and combine with other factors such as environment. This is what the HapMap project is about, and Donnelly is at the forefront of looking at the variation of genes within the human species - to try and find out what is really relevant to disease. Donnelly has also developed a fundamental framework for analysing genome data that is leading to conclusions about human evolution, and even early human history, by tracking the migrations of populations through the variations in their genes.

David Altshuler, assistant professor of genetics and of medicine at Harvard Medical School, and co-chair with Donnelly and one other of the data analysis group on the HapMap project, said: 'He is one of the clear world leaders in the field of using mathematical and computational methods to analyse DNA variation data. He has made huge contributions himself and he has trained a group of young people around the world who are, themselves, now becoming world leaders. It is impossible to define the best, but there is certainly no one better. He comes from the theoretical end of the field, but in recent years he has become more involved in the data-collection part of the field and that is relatively unusual.

'His most important contributions are firstly the people who have emerged from his lab, who are the brightest young people in the field of computational analysis of genetic variation. Secondly, his lab has pioneered models to extract information about the underlying biology of variation and mutation and about the history of mankind. They are at the cutting edge of developing the methods used to deal with real data.

'A great collaborator'
'Peter has no formal biology training, but he has teamed up people who have that expertise. Peter is an extremely nice person, is very reasonable, and has excellent judgment, and so he is a great collaborator. If he was not such a good collaborator and good to work with, he might have a harder time with the things he does not have formal training in.

'What makes him really unique is that he engaged with biological problems and biologists. He is a fun guy to work with; he has a twinkle in his eye and a great enthusiasm for opportunities and ideas. He has a tremendous presence about him, he speaks with authority. There is a lot of culture-clash, because so many different kinds of people are involved in the field - but Peter is one of the few people in the field who everyone likes and respects.'

Donnelly was born in Brisbane, in the tropical state of Queensland, Australia. He was interested in mathematics from an early age but, like most Australians, was also a keen sportsman, playing cricket and rugby. He showed his promise early and went to University in Brisbane a year earlier than most of his contemporaries. He studied mathematics and physics at first, later dropping physics. He had grappled with the idea of studying law, as he was a keen member of the school debating team. In later years, he would combine the two as an expert witness in criminal cases involving the use of DNA profiles as evidence. Between his third and fourth year, he got a vacation scholarship to study at the Australian National University in Canberra, where he mixed with researchers and students from all over the country. He was working with the statistics research group of Professor Ted Hannan. Donnelly said: 'It opened my eyes to how exciting and challenging the subject could be. I went back to do my fourth year with great enthusiasm for the subject.'

After graduating with first-class honours, he applied for and was awarded a Rhodes Scholarship to further his studies at Oxford. Only one such scholarship is awarded in each Australian State.

The 'Academic Brat Pack'
Life at Oxford was great fun for Donnelly. Like most Australians he had gone to his local university and had lived at home while he studied. Oxford was not only his first experience of living away from home, but was also overseas. He was slightly younger than most research students, and admits that he did not work as hard as he should have, and enjoyed the life more associated with undergraduates, playing lots of rugby and cricket. His thesis was on stochastic processes and he was even at this stage beginning to see ways of applying them to genetics, although the subject was not as advanced in those days.

After gaining his doctorate, he looked around for a post-doc position. He looked around the UK but found few jobs on offer, and was encouraged by a friend from ANU to apply to the University of Michigan, where he stayed for just over a year. He had by then married an English woman, so he looked again for a position in the UK, eventually finding a job at the University of Wales in Swansea. He moved then to University College London and, in 1988, at the age of 29, became the UK's youngest professor, at Queen Mary and Westfield College. He featured in newspaper stories about the 'Academic Brat Pack'.

His research continued around the statistics of genetics, which led to him being called as an expert witness in criminal cases. Mostly, they were appeal cases where he was called to put into context claims about the uncertainty figures quoted for DNA evidence.

Donnelly says he was really happy in London. He spent a lot of time travelling and collaborating with people in the US. He had to think very hard when he was approached by the University of Chicago, but he got a great offer of a joint position in the department of Statistics and department of Ecology and Evolution. This allowed him to assemble an inter-disciplinary research team. When he got there, he has glad he did and remembers it as a great place to work. His department expanded from eight to 20 academic staff, and research funding flowed.

Oxford calling
So, two years later, when he was approached by Oxford, he had an even harder decision to make. It was a prestigious position that would probably not come up again for another 20 years. In the end, personal factors played a part. He said: 'By that time I had two children, with a third on the way, and we preferred for them to grow up in Oxford.'

Many would be nervous of turning their backs on the grant funding that is found in the US, but Donnelly has not found any problems. He said: 'The stereotype is that there is more money available in the US, but I have found the university to be very supportive and they have let me expand the department.'

Much of his research has centred on genetic variations within human populations. He has found that there is much more variation in the populations of certain African countries than in Europe, suggesting a genetic 'bottleneck' as humans began to migrate north from Africa. Variation can not only tell you things about the history of human migration, it can also tell you things about the processes involved in variation by studying modern populations. Human reproduction makes one copy of a genome sequence from two copies, one contributed by each parent. He has been able to study the process of recombination to find out what parts of the genome are chosen to make up the one copy.

The next stage in variation research is to look at disease. Although a very few conditions, for example Huntington's Disease, are associated with a single gene that is either passed on or not, most other conditions have a more complex relationship with genetics. Environment and other factors also play a part. One goal is to be able to spot variations which might lead to somebody being at greater risk of developing type 2 diabetes or heart disease. This work has put him at the centre of some of the highest-profile research in the world, as HapMap is sometimes described as the successor to the Human Genome project. He is being called upon to create the frameworks that will be used to turn the huge amounts of data being produced into a form that biologists can use to learn more about the way genes actually work.

As one might expect of an Oxford Don, he still finds time for a little relaxation. He has found that, with cricket, it is always possible to find your own level. One assumes that, being an Australian, playing cricket is somewhere in his genes. Maybe, one day, he will find out exactly where.

Peter Donnelly CV

Education
1976-79
University of Queensland, BSc Mathematics
1980-1983
University of Oxford, DPhil, Mathematics

Employment
1983-84
University of Michigan, visiting assistant professor, Department of Statistics
1984-85
University of Wales, Research fellow, Department of Mathematics and Computer Science
1985-88
University College, London, Lecturer, Department of Statistical Science
1988-94
Queen Mary and Westfield College, University of London, Professor of Mathematical Statistics and Operational Research
1994-96
University of Chicago, Professor, Department of Statistics and Department of Ecology and Evolution
1996-date
University of Oxford, Professor and Head of Department of Statistics.