Grant: Discovery Grants Program - Individual

The coalescent process, a stochastic process used to simulate population genealogies, has attracted much attention recently, mostly due to the numerous applications in human genetics research. We have developed a method, which uses the coalescent process with recombination (Larribe et al. 2002, Larribe 2004) to estimate the location of a genetic mutation influencing a disease, e.g. to do genetic fine mapping. After having recently improved the stability of the estimates using a composite likelihood strategy (Larribe and Morgan, 2004), we propose to develop the method in the context of complex genetic disease, which is of major interest in human genetics today. The goal is to model parameters of incomplete penetrance and phenocopy rate, and treat them as nuisance parameters in the estimation, in a similar way as we do with unknown genealogies. Morover, we would like to develop the method so that it is capable of estimating the position of two mutations simultaneously. Finally, we want to correct for the ascertainment biases (unequal sampling of cases and controls), by modelling the frequency of the disease mutation over time, from today back until the observed sequences have a common ancestor.