|
Activity Number:
|
81
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, July 30, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #307993 |
|
Title:
|
Using Sequence Information to Predict Gene Regulation
|
|
Author(s):
|
Jun S. Liu and Qing Zhou*+
|
|
Companies:
|
Harvard University and University of California, Los Angeles
|
|
Address:
|
Depatment of Statistics, Los Angeles, CA, 90095-1554,
|
|
Keywords:
|
Regression trees ; Gene regulation ; Bayesian model average ; boosting ; regression ; MCMC
|
|
Abstract:
|
Understanding how genes are regulated in various circumstances is a central problem in molecular biology. The adoption of large-scale biological data generation techniques such as the expression microarrays has enabled researchers to tackle the gene regulation problem in a global way. We present a method based on the Bayesian Additive Regression Trees (BART) developed by Chipman et al. (2006) for extracting sequence features to predict gene expression or enrichment values. We show that BART significantly outperformed the neural network and our earlier stepwise linear regression approaches, for real datasets including the human Oct4 and Sox2 ChIP-chip datasets and yeast amino acid starvation dataset. The variables selected by BART are also of significant biological significance.
|
