|
Activity Number:
|
520
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #304684 |
|
Title:
|
A Score Test Based on Modified Genotype Data
|
|
Author(s):
|
Renfang Jiang*+ and Yilin Dai and Jianping Dong
|
|
Companies:
|
Michigan Technological University and Michigan Technological University and Michigan Technological University
|
|
Address:
|
Department of mathematical Sciences, Houghton, MI, 49931,
|
|
Keywords:
|
genome-wide association studies ; multilocus association tests ; wavelet transform ; empirical Bayes threshold ; linkage disequilibrium
|
|
Abstract:
|
Reducing the number of degrees of freedom by suppressing noise from data is a key to increasing the power of multi-locus association tests. One difficulty is to decide how much noise to suppress. We developed a score test based on wavelet transformation. Wavelet transformation is designed to deal with unsmooth noisy signals. Using an empirical Bayesian thresholding, the proposed test suppresses more data if many single nucleotide polymorphisms (SNP) under consideration are unrelated to the disease, and it suppresses less if most SNPs are associated with the disease. Simulations show that the proposed test has a higher power than the commonly used single marker methods, the multiple logistic regression, and the principal components regression method for various LD structures. The proposed test can be applied to both qualitative and quantitative traits.
|
