McMaster University

Graduate Program in Statistics



STATISTICS SEMINAR



SPEAKER:
Shelley Bull
Samuel Lunenfeld Research Institute,
Mount Sinai Hospital, Toronto
Date :Wednesday March 24, 2004.
Time : 3:00pm
Address Hamilton Hall
Room: 217
TITLE:
Regression Models for Allele Sharing in Affected Relatives
ABSTRACT:
Modelling of variation in identical-by-descent (IBD) allele sharing using covariates can be useful to increase power to detect linkage, identify covariate-defined subgroups that are linked to particular marker regions, and improve the design of subsequent studies to localize genes and characterize their effects in combination with other genes and/or environmental factors. In this talk, we highlight and explore issues that arise in studies of families with affected relatives. We have extended the linear and exponential linkage likelihoods described by Kong and Cox [1997] to model variation in non-parametric linkage (NPL) scores among covariate-defined groups of families with affected relatives. We proposed a likelihood ratio (LR) test for the covariate contribution and compared its performance to a simple t-statistic for group mean NPL differences in the case of a binary covariate. In simulation studies of locus heterogeneity in families with affected siblings or affected cousins, we show how the distribution of LR test statistics depends on the extent of linkage, particularly in the presence of constraints on the parameters. On the other hand, the distribution of the t-statistics may be biased by differences between groups in information content. We recommend that constraints on the parameters be applied with caution, and interpretation of covariate effects in IBD allele-sharing models made with care.
About the Speaker
Shelley Bull is Senior Scientist in the Samuel Lunenfeld Research Institute of Mount Sinai Hospital, Professor in the Department of Public Health Sciences at the University of Toronto, and a Senior Investigator of the Canadian Institutes of Health Research. Her research program encompasses independent research in biostatistics and collaborative research with investigators in epidemiology and molecular biology. Her undergraduate and master's studies were in mathematics and statistics at the University of Waterloo, with doctoral and post-doctoral training in biostatistics and epidemiology at the University of Western Ontario. She is currently project leader of the MITACS project in "Statistical genetic modelling and analysis of complex traits".
References
Some suggestions for background reading are:


Department of Mathematics and Statistics
Graduate Program in Statistics

This page is maintained by Angelo Canty,
Last updated on March 18, 2004