|
Activity Number:
|
548
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 2, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #308175 |
|
Title:
|
Application of Johnson System of Distribution in Gene Selection
|
|
Author(s):
|
Florence George*+ and Kandethody Ramachandran
|
|
Companies:
|
University of South Florida and University of South Florida
|
|
Address:
|
10716 Tavistock Dr, Tampa, FL, 33626,
|
|
Keywords:
|
differentially expressed genes ; Johnson system of distributions
|
|
Abstract:
|
Microarrays have become increasingly common in biological and medical research. They enable the simultaneous study of thousands of genes and provide gene expression information on a whole genome level. A major goal of microarray experiments is to determine which genes are differentially expressed between samples. A new approach using the Johnson's system of distributions is proposed in this paper to make inference concerning which genes are differentially expressed, in which no specific parametric distribution is assumed for the gene expression levels. Here the null distribution of the summary statistic is estimated empirically using Johnson's system of distributions. The simulation results show that the new approach has a higher power over the typical traditional methods.
|
