EPSRC Reference: 
GR/R91724/01 
Title: 
Model Choice and Robustness in Population genetics 
Principal Investigator: 
Fearnhead, Professor P 
Other Investigators: 

Researcher CoInvestigators: 

Project Partners: 

Department: 
Mathematics and Statistics 
Organisation: 
Lancaster University 
Scheme: 
Fast Stream 
Starts: 
01 October 2002 
Ends: 
30 September 2005 
Value (£): 
57,792

EPSRC Research Topic Classifications: 
Bioinformatics 
Population Ecology 
Statistics & Appl. Probability 


EPSRC Industrial Sector Classifications: 
No relevance to Underpinning Sectors 


Related Grants: 

Panel History: 

Summary on Grant Application Form 
W e will consider the analysis of infintesites population data (that is sequence data that is consistent with no recombination and no repeat mutations). W e wil develop methods that can analyze the data under a large class of possible demographic models, that enable the evidence for any demographic model to be calculated (thereby aiding model choice), and that can efficiently analyze a single data set under a wide variety of models. This will be achieved by first developing MCMC algorithms that can analyze infinite sites data under an uninformative prior. This MCMC agorithm will be efficient as the proposal density of the MCMC algorithm will be designed to take account of the characteristics of population data for example the dependence (and conditional independence) structure of the branch lengths in the unobserved genealogy. The MCMC algorithm will form the basis of the computational algorithms that we will develop to analyze the infinite sites data under a wide variety of demographic models. One of the proposed algorithms is based on the idea of Importance sampling. Given a sample from our MCMC algorithm (that is a sample from the posterior distribution of genealogies, under our uninformative prior), we can obtain a weighted sample from the posterior distribution under a different model by assigning each genealogy a weight proportional to the prior probability of the genealogy under the model of interest. This method is firstly efficient, as the calculation of the importance sampling weights is computationally inexpensive (and thus a single data set can be easily analyzed under various different models). Secondly, it enables the evidence for each model to be calculated.

Key Findings 
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk

Potential use in nonacademic contexts 
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk

Impacts 
Description 
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk 
Summary 

Date Materialised 


Sectors submitted by the Researcher 
This information can now be found on Gateway to Research (GtR) http://gtr.rcuk.ac.uk

Project URL: 

Further Information: 

Organisation Website: 
http://www.lancs.ac.uk 