|
Activity Number:
|
510
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 10, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #306145 |
|
Title:
|
A Valid False Discovery Rate Procedure in Presence of Biased Null P-values
|
|
Author(s):
|
Hoa Phuong Nguyen*+
|
|
Companies:
|
Johnson & Johnson
|
|
Address:
|
ALZA Plaza, 1950 Charleston Road, Mountain View, CA, 94043,
|
|
Keywords:
|
FDR ; SNPS ; microarray ; genome wide disease association
|
|
Abstract:
|
In recent years, genome wide disease association effort has been focused on (1)identifying gene expression differentials using microarray technology and (2)searching through millions of single nucleotide polymorphisms (SNPs) for liability alleles affecting complex diseases. Due to a large number of hypotheses tested to assess association, the FDR procedure introduced by Benjamini and Hochberg (BH) has become mainstream in controlling the combined type-I error rate. The authors showed validity of the BH procedure assuming uniform p-values of null hypotheses. In this talk, I discuss genome wide disease association scenarios in which biased (non-uniform) null p-values can arise and how biased p-values degrade performance of the BH procedure. I propose the tcFDR procedure which improves upon the performance of the BH procedure in presence of biased null p-values.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2006 program |
