|
Activity Number:
|
377
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #308978 |
|
Title:
|
Weighted Distance Measures for Metabolomic Data
|
|
Author(s):
|
Philip Dixon*+
|
|
Companies:
|
Iowa State University
|
|
Address:
|
120 Snedecor Hall, Ames, IA, 50011-1210,
|
|
Keywords:
|
cluster analysis ; PCA
|
|
Abstract:
|
A typical metabolomic study measures the concentration of a very large number of cellular metabolites in a relatively small number of samples. Biologists frequently analyze metabolomic data using a cluster analysis based on Euclidian distance between samples or a Principal Components Analysis, which is indirectly based on Euclidian distance. We examine patterns of variation between replicate samples in two data sets and find that the s.d. is approximately proportional to the mean. Based on this, we develop three weighted measures of distance that better differentiate treatments. The Canberra distance is equivalent to one of these weighted measures. A modification of the weights provides a reasonable analysis of samples with below detection-limit values.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2007 program |
