To better understand multifactorial diseases such as cancer, diabetes and cardiovascular diseases, and to ultimately better target treatments to individuals, researchers are using new biotechnologies that measure genetic code at extremely high resolution as well as downstream functional mechanisms essential to the maintenance of human health, and study designs that combine extensive questionnaires, genotyping and biobanks. However, the amount and diversity of information collected render their analysis difficult and statisticians are faced with the challenge of developing efficient dimension reduction approaches that can discover important predictors and patterns among a vast array of possibilities. Our programme proposes to develop a range of improved statistical techniques and algorithms for finding important combinations of features in large genetic and genomics datasets that characterise or predict health outcomes and for carrying out integrative analyses to characterise heterogeneous disease processes. The new methods will be accompanied by the development of freely available software and will be used in a number of collaborative projects to improve understanding of the regulation of genes and immunological response, to study gene-environment interactions and to develop biomarker-based prognostic signatures.
Leads the Big data and Public Health topic with Nigel Collier
