|
Activity Number:
|
510
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 10, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #305565 |
|
Title:
|
Validation of Biomarkers Identified by Gene Expression Profiles
|
|
Author(s):
|
Boris Zaslavsky*+ and Jing Han and Jawahar Tiwari and Raj K. Puri
|
|
Companies:
|
U.S. Food and Drug Administration and U.S. Food and Drug Administration and U.S. Food and Drug Administration and U.S. Food and Drug Administration
|
|
Address:
|
1401 Rockville Pike, Rockville, MD, 20852,
|
|
Keywords:
|
logistic regression ; gene expression ; perturbation ; leave-one-out cross-validation
|
|
Abstract:
|
Two hundred ninety three human embryonic kidney (HEK) cells were cultured under different confluence states and RNA isolated and labeled targets were hybridized with high quality ~10K cDNA microarrays. The logistic regression method of SAS software was applied to identify a set of biomarkers by gene expression profiles. Two-group classification (i.e., over confluence vs. 90% confluence) was used to select the subset of differentially expressed genes. To evaluate the robustness or stability of the logistic classifier, we generated simulated datasets by perturbation of the raw dataset with various error levels. We also used leave-one-out cross-validation procedure and perturbed datasets to estimate the subset of most informative genes. Results confirmed that simulated datasets based on the perturbation of the raw data can be used for the validation of biomarkers.
|
