Modelling Genes
by
Peter Donnelly
University of Oxford

Following on from the various genome projects, modern experimental techniques in genetics are giving rise to vast amounts of data and challenging statistical problems. Growing amounts of this data document genetic variation, at the level of DNA sequences, between individuals within populations. Such data has the potential to shed light on (i) the demographic histories of the populations from which the individuals are sampled, (ii) the underlying evolutionary mechanisms, and (iii) the genetic basis of common diseases (notably in humans). But interpretation of the data is far from straightforward. It arises from the complex interactions of evolutionary and genetic forces during the populations' evolution. Recent progress in the field has combined relatively sophisticated stochastic models for population evolution with tools from computationally intensive statistical methodology. The talk aims to convey the interest in the science and to give a flavour of the some of the statistical problems to which it leads.

Date received: June 2, 2002