|
Activity Number:
|
470
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #304702 |
|
Title:
|
Score Statistics for Family-Based Genetic Association Studies of Quantitative Traits
|
|
Author(s):
|
Samsiddhi Bhattacharjee*+ and Eleanor Feingold
|
|
Companies:
|
National Cancer Institute and University of Pittsburgh
|
|
Address:
|
, , ,
|
|
Keywords:
|
Genetics ; Association ; Score Test ; Population Stratification ; Family-based ; Pedigree Founders
|
|
Abstract:
|
Family-based tests of genetic association protect from spurious associations by ignoring and/or conditioning on certain data for the pedigree founders. Hence, these tests are often considerably less powerful than population-based tests that use all of the available information but are sensitive to stratification. We present a unified likelihood for quantitative traits in families and derive novel score statistics that provide a continuum of options between purely population-based and traditional family-based tests. They incorporate founder phenotypes and derive additional power from founder genotype-phenotype correlation and founder-non-founder environmental correlation using certain assumptions about the nature of stratification. We derive closed form expressions for null means and variances of these statistics and compare their performance to the standard procedures using simulations.
|
