FACES OF MATHEMATICS

Peter Donnelly is Professor of Statistical Science at Oxford University. His main research interests are in the application of probability and statistics to genetics, where theoretical and practical advances in probability modelling are applied to the understanding of evolutionary history and the structure of the human genome. His work combines stochastic processes and computationally intensive statistics, and involves extensive collaborations with biologists to develop new methodologies in genetics.

"There are really nice abstract mathematical problems which happen to have this genetic application. But these days my primary motivation is science. We really can contribute at a fundamental level to solving these scientific problems.

"It's a real challenge to find ways of using all the information in the data. We've made progress through fairly sophisticated mathematical analysis of the probability models. It turns out that if you do that carefully, you can end up with enormous practical advantages. It is a nice reward: thinking hard about the stochastic models is really helpful in analysing the data.

"Having done the modelling, you get the insight and the understanding is there. Usually, the naive interpretation of genetic data is misleading because of the subtleties involved in the processes generating the data. That is why it is worth thinking about mathematically."