|
Activity Number:
|
387
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #310040 |
|
Title:
|
Family-Based Case-Control Studies
|
|
Author(s):
|
Li Zhang*+ and Bhramar Mukherjee
|
|
Companies:
|
The Cleveland Clinic and University of Michigan
|
|
Address:
|
6809 Mayfiled Road, Cleveland, OH, 44124,
|
|
Keywords:
|
Dirichlet Process ; hierarchical pedigree ; conditional likelihood ; mixed effects model ; random effects model
|
|
Abstract:
|
As a compromise between linkage studies and population-based case-control studies, family-based association designs have received great attention recently due to their potentially higher power to identify complex disease genes and their robustness in the presence of population substructure. Based on a two-level mixed effects model which allows to estimate environmental effects while accounting for varying genetic correlations among family members and adjusting for ascertainment by conditioning on the number of cases in the family, we propose a full Bayesian alternative to build in a hierarchical pedigree structure and assuming priors on the random effects which offers a more appealing alternative. We provide a general framework for Bayesian analysis for the random effects model where a nonparametric Dirichlet Process prior is specified for the random effects.
|
