|
Activity Number:
|
379
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
WNAR
|
| Abstract - #304812 |
|
Title:
|
Using Exon Microarrays to Predict Breast Cancer Occurrence
|
|
Author(s):
|
William E. Johnson*+ and Ying Sun and Andrea Bild
|
|
Companies:
|
Brigham Young University and The University of Utah and The University of Utah
|
|
Address:
|
Department of Statistics, Provo, UT, 84602,
|
|
Keywords:
|
Exon microarray ; Breast Cancer ; Normalization ; Mixture model ; Bayesian Hierarchical Model
|
|
Abstract:
|
Exon microarrays are valuable tools for measuring gene expression and detecting alternatively spliced exons. In this research we present complete analysis method for analyzing Affymetrix exon arrays. We apply a novel normalization method that uses a mixture model to identify and filter probe/array effects while leaving biological signal untouched. This substantially increases the signal-to-noise ratio in the data, allowing for clearer identification of interesting biological features. Additionally, we present a Bayesian Hierarchical model that allows for the identification of differentially expressed gens, spliced exons and differentially spliced exons. The model is flexible and can be applied to individual or grouped samples. We apply our method with great success to a breast cancer data set with the goal of predicting whether an individual will develop breast cancer in her lifetime.
|
