|
Activity Number:
|
365
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #301662 |
|
Title:
|
Statistical Analysis of Censored Proteomic Expression Data
|
|
Author(s):
|
Shiying Wu*+ and James Stephenson and Carol Whisnant and Diane Wagener
|
|
Companies:
|
RTI International and RTI International and RTI International and RTI International
|
|
Address:
|
3040 Cornwallis Road, Research Triangle Park, NC, 27709,
|
|
Keywords:
|
censored data ; measurement error ; partial correlation ; proteomic expression ; regression ; typhoid vaccination
|
|
Abstract:
|
The association between immune response to typhoid vaccination and the expression levels of a set of defensins was investigated in this study. Defensins are a major class of antimicrobial peptides in mammals. The expression measurement of defensins was often heavily censored due to the detection and quantitation limits of the bioassay, and the accuracy of the immune response status was also limited by the accuracy of the bioassay. In addition, the data were not normally-distributed. To cope with these problems, the nonparametric test for first-order partial correlation with censored data (Akritas and Siebert, 1996) was first generalized to the higher order case and then applied. A censored regression model with defensin concentration as dependent variable and lognormal model errors was also applied for comparison. Simulation was conducted to evaluation the two approaches. The result is reported in this presentation.
|
