Semi-nonparametric Model for Cell Cycle Microarray Data

• Activity Number: 415
• Type: Contributed
• Date/Time: Thursday, August 12, 2004 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Guei-Feng Tsai*+ and Annie Qu
• Companies: Oregon State University and Oregon State University
• Keywords: microarray ; cluster ; nonparametric ; quadratic inference function

Abstract

A new approach is developed to deal with cell cycle microarray data. There are two kinds of correlations in cell cycle data. Measurements are certainly correlated within a gene where it is measured over cycles, and measurements could also be correlated between genes, since some genes are biologically related and regulate the same phenotypical characteristics. The proposed procedure combines cluster data analysis, the quadratic inference function method and nonparametric techniques for complex high-dimensional longitudinal data. We first perform clustering data analysis to classify genes with similar cell cycle patterns into the same class, or into a class with no cell cycle phenomena at all. We use genes within the same cluster as replicates to develop nonparametric models. To incorporate correlation of longitudinal measurements, the quadratic inference function method is applied. This also allows us to perform a goodness-of-fit test for testing whether the coefficients are time varying. This leads us to the determination of whether certain genes regulate cell cycles.