Fall Seminar Series - November 20, 2003
University of Minnesota
School of Statistics
College of Liberal Arts
Fine-scale mapping of disease genes based on linkage disequilibrium
Na Li
Division of Biostatistics
University of Minnesota
Thursday, November
20, 2003
4:00 PM, 115
Ford Hall
Minneapolis, East
Bank Campus
Social at 3:30
PM, 300
Ford Hall
Abstract
Genetic
mapping is based on the fundamental idea that the probability of recombination
between markers is higher for markers that are further apart, and lower for
those that are closer. When markers are very close, the probability of observing
any recombination within a family (even a large one) becomes extremely small,
which limits the resolution of pedigree based linkage analysis. For fine-scale
localization, a population based approach (e.g., a case-control design) is
not only more powerful but sometimes necessary. Apparently unrelated individuals
are in fact related though distant co-ancestry and the effect of recombinations
through many generations is reflected on the patterns of linkage disequilibrium
(association of marker alleles at different loci in the population) in the
current generation. In this talk we will first describe the basic idea behind
current approaches to the fine-mapping problem and discuss their limitations.
Then we will introduce a new method based on the Product of Approximate Conditionals
(PAC) model (Li and Matthew, 2003). Some simulation results are shown.
This is joint work with Matthew Stephens, Department of Statistics, University
of Washington.
Seminar jointly sponsored by School of Statistics and Division of Biostatistics