|
Activity Number:
|
155
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 7, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #306923 |
|
Title:
|
Cluster Analysis for Gene Expression Data with Liquid Association Structure
|
|
Author(s):
|
Yijing Shen*+ and Ker-Chau Li and Shinsheng Yuan
|
|
Companies:
|
University of California, Los Angeles and University of California, Los Angeles and University of California, Los Angeles
|
|
Address:
|
11140 Rose Ave., Los Angeles, CA, 90034,
|
|
Keywords:
|
genes ; clustering ; liquid association ; k-means ; EM algorithm
|
|
Abstract:
|
An important goal in gene clustering analysis is to identify functionally related gene groups that could be used in functional annotation for new or uncharacterized genes. We analyze small gene sets selected from the Yeast genome by the method of Liquid Association (Li, 2002)-the correlation between the expression levels of such gene pairs depends on the change of a relevant external variable. Traditional approaches may not always provide the best results because functionally related genes may only be co-expressed under a subset of conditions. For such structured datasets, expression profile of the external variable is an important factor in guiding the clustering. By employing the external information we develop a new clustering algorithm that optimally selects two subsets of conditions for generating best clustering; then we refine the clustering result through the use of Gene Ontology.
|
