|
Activity Number:
|
81
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 12, 2002 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biometrics Section*
|
| Abstract - #301760 |
|
Title:
|
Interpreting and Estimating the P-Values of a Microarray Experiment: a Comparative Study
|
|
Author(s):
|
Manisha Desai*+ and Ying-Kuen Cheung
|
|
Affiliation(s):
|
Columbia University and Columbia University
|
|
Address:
|
630 West 168th Street, PH18W, New York, New York, 10032, USA
|
|
Keywords:
|
microarrays ; multiple comparisons ; mixture models
|
|
Abstract:
|
Microarrays are a new technology that allow the simultaneous monitoring of expression levels for thousands of genes. Scientists are interested in knowing whether the expression level of a particular gene significantly differs among sample types or groups. A consequence of this is that hypotheses are posed for each of thousands of genes included in an experiment, leading to a large multiple comparisons problem. Dudoit et al. (2001) apply a method developed by Westfall and Young (1993) for analyzing microarray data in order to obtain p-values adjusted for multiple comparisons. A disadvantage of this method is that it is conservative, leading to a high false negative rate. A further limitation is the interpretation of the adjusted p-value, which is conditional on all the null hypotheses holding; no interpretation is provided if one of the hypotheses do not hold. We approach this problem by modelling the p-values as a mixture of two distributions, where the goal is to estimate the probability that a gene-specific p-value is significant. We compare the two approaches via a simulation study which examines the size and power of the methods under various scenarios.
|
- 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 2002 program |
