|
Activity Number:
|
154
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #308579 |
|
Title:
|
A Semiparametric Mixture Model for Identifying Cis-Acting Regulatory Elements
|
|
Author(s):
|
Gregory Hather*+ and Terence Speed and Mary Wildermuth
|
|
Companies:
|
University of California, Berkeley and University of California, Berkeley and University of California, Berkeley
|
|
Address:
|
367 Evans Hall, Berkeley, CA, 94720-3860,
|
|
Keywords:
|
cis-acting regulatory elements ; binding sites ; transcriptional regulation ; sequence analysis
|
|
Abstract:
|
Traditional computational methods for identifying cis-acting regulatory elements do not effectively integrate gene expression data into the analysis. We present a model based on the work of Wang et al. (2005) that relates these two types of data. One input to our model is a list of short DNA segments that fall within the promoter region of a gene and that match a user-specified core motif. The other input is the gene expression data. Each DNA segment is assumed to be either active or inactive in terms of its role in regulating gene expression. Any gene with at least one active segment nearby is assumed to be active. Activity is a binary hidden variable that depends upon the sequence data and that influences the expression data. The model can be used to estimate the probability that a given DNA segment is active under the conditions for which the gene expression was measured.
|
