MRC Co-operative Group in "The Genetic Basis of Neuropsychiatric and
Neurodegenerative Disorders" and the Biostatistics & Bioinformatics Unit,
Department of Psychological Medicine, Wales College of Medicine, Cardiff
University, UK.
Applications are invited for the following PhD studentship funded by MRC and
Departmental funds, to start October 1st 2005
Efficient methods for incorporating covariates in model-free genetic linkage
analyses.
Prof P Holmans and Dr M Hamshere. ([log in to unmask])
Project Description
The inclusion of covariates in model-free linkage analysis is a promising way to
increase the power of linkage studies in complex traits. A major advantage of
such an approach is its flexibility: it can combine different kinds of
covariate, for example, phenotypic information (e.g. disease severity, age at
onset or clinical subtypes) or known risk factors (genetic or environmental).
A number of methods have been proposed to perform model-free linkage analysis
with covariates on affected sibling pairs, but as yet there have been few
comparisons of their power. A major thrust of this project, therefore, will be
to perform a comprehensive and systematic study of the relative power of the
various methods under a number of models for multilocus genetic susceptibility
and environmental effects. The effects of incomplete linkage information from
the genetic markers will also be studied. This will enable conclusions to be
drawn both regarding the situations in which using covariates will increase
power to detect linkage and the optimal analysis methods in these situations.
The project will also involve extending the methods from affected sibling pairs
to more general pedigrees, investigating the statistical properties of such
extensions, and determining the most effective types of pedigree to use for a
given type of covariate or objective. At present, the most effective way of
coding covariates is unclear, as is the most powerful choice of test statistic.
The project will investigate both of these issues. Finally, the investigation of
the properties of the various methods will lead to the development of novel,
more powerful, analysis methods, and this will be an important part of the
project. The project will involve intensive computer simulation studies,
together with applications to real datasets provided by the supervisor's
collaborators. Some extensive datasets are already available (in schizophrenia,
bipolar disorder, late-onset Alzheimer's disease and recurrent early-onset
depression), and it is envisaged that others will become available during the
course of the project.
The Department contains a wide range of scientific expertise relevant to genetic
and other aspects of psychiatric disorders including phenotype definition,
epidemiology, statistical genetics, molecular genetics, cell biology, proteomics
and transcriptomics, biochemistry, neural transplantation, stem cell biology and
the analysis of animal models of neuropsychiatric disorder. The Department can
also provide advanced training in bioinformatics, including exploitation of
databases and software for handling extremely large amounts of data. More
information can be found at
http://www.cardiff.ac.uk/medicine/psychological_medicine/research/index.htm and
http://bbu.uwcm.ac.uk/index.htm
Candidates should have, or expect to obtain, a First or high Upper Second Class
Honours degree in the biological, psychological, computational or mathematical
sciences. Some computer programming ability and knowledge of statistics is
essential. Knowledge of genetics is desirable. For further details of the
project or informal enquiries email Peter Holmans ([log in to unmask])
Applications should be made by sending a CV and covering letter to Sue Cody,
Psychological Medicine, Cardiff University, Heath Park, Cardiff, CF14 4XN (or by
email to [log in to unmask])
Closing date for receipt of applications is 30th June 2005
|