|
Activity Number:
|
401
|
|
Type:
|
Invited
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
IMS
|
| Abstract - #303082 |
|
Title:
|
Statistically Based Mining of Copy Number and Correlation Matrices
|
|
Author(s):
|
Andrew Nobel*+
|
|
Companies:
|
The University of North Carolina at Chapel Hill
|
|
Address:
|
Department of Statistics and OR, Chapel Hill, NC, 27599-3260,
|
|
Keywords:
|
data mining ; high dimensional data ; transcriptome map ; comparative genome hybridization ; prediction
|
|
Abstract:
|
The large average submatrix (LAS) algorithm is a statistically based data mining procedure that identifies significant sample-variable associations in high dimensional data sets. The LAS algorithm has previously been applied to gene expression studies of breast and lung cancer. In this talk we will discuss several ongoing extensions of the basic LAS procedure. The first extension concerns application of LAS, with modifications of its score function, to the analysis of high-throughput array CGH data that measures copy number variation in the human genome. The second extension is to the mining of correlation matrices arising in experiments where multiple measurement technologies are applied to a common set of samples. Of particular interest is the mining of transcriptome maps that arise in expression QTL analyses.
|
