|
|
KATHLEEN KERR |
Statistical Design and Analysis of Expression Microarray
Experiments |
Keywords: Analysis of variance, experimental design, block design,
gene expression, shrinkage estimators, multiple testing |
This research will establish a solid
statistical framework for the design and analysis of gene expression microarray
experiments. |
Microarrays ultimately produce a set of data intended
to answer specific biological questions. We propose a class of ANOVA models
to analyze this data. Variations of these models will be considered, such
as robust estimation, non-parametric inference, and including random effects
and employing shrinkage estimators. We will develop experimental designs
to facilitate our analysis and increase precision and confidence in the
results. Although our designs will be oriented towards this method of analysis,
their value will not depend on a specific model. In addition, we will seek
designs that are robust to potential problems with the experiment such
as loss of an array or contaminated spots. We will employ our design and
analysis methodologies in experiments with collaborators at The Jackson
Laboratory. One application of this work is a study of gene expression
differences among strains of mice with differential susceptibility to type
II diabetes. A second application is to identify genes underlying differential
susceptibility to gallbladder disease. |
|
|
