Analyzing Temporal Microarray Gene Expression Data with Replicates

• Activity Number: 258
• Type: Contributed
• Date/Time: Tuesday, August 13, 2002 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section*
• Author(s): Ronghui Xu*+
• Affiliation(s): Harvard School of Public Health
• Keywords: temporal microarray gene expression ; replicated data ; clustering ; permutation

Abstract

In a biological experiment to characterize the signaling to B cells through beta-1 integrins in order to study its effect on B cell survival, Fibronectin and Poly-L-lysine were used as stimulant and control on two leukemia B cell lines for various lengths of time. Microarray data were collected to study the temporal profiles of gene expression. In this talk, we describe the various statistical methods applied to the temporal expression data in order to select interesting genes, which led to the identification of genes that are affected by adhesion in human B cells that regulate apoptosis. The gene selection methods include pointwise comparisons of gene expressions and nonlinear modelling of expression data over time, using both parametric and permutation approaches. Gene clustering also help selection of genes, and we indicate the limitation of existing clustering methods when applied to replicated data.