Gene Expression Arrays

Overview
Analysis of variance for expression data
Experimental design for microarrays
Bootstrapping cluster Analysis
Design and analysis of microarrays
Statistical analysis of a gene expression microarray experiment with replication
Analysis of a designed microarray experiment

Experimental design for gene expression microarrays

M. Kathleen Kerr and Gary Churchill

Abstract:We examine experimental design issues arising with gene expression microarray technology. Microarray experiments have multiple sources of variation, and experimental plans should ensure that effects of interest are not confounded with ancillary effects. A commonly-used design is shown to violate this principle and to be generally inefficient. We explore the connection between microarray designs and classical block design and use a family of ANOVA models as a guide to choosing a design. We combine principles of good design and A-optimality to give a general set of recommendations for design with microarrays. These recommendations are illustrated in detail for one kind of experimental objective, where we also give the results of a computer search for good designs. Manuscript

The following files comprise a catalog of designs referred to in ``Experimental Design for Gene Expression Microarrays.'' The files contain A-optimal designs for block size two, number of varieties v=6,7,8,9,10, and number of blocks between v and v(v-1)/2. Additional files gives A-optimal designs for v=11,12,13 varieties and small numbers of blocks. The A-optimal designs were found by (1) generating all non-isomorphic connected graphs on v vertices using Brendan McKay's, makeg program, and (2) comparing all designs of the same size on the basis of A-optimality.

The files contain lists of the two best (in terms of A-optimality) designs for given numbers of varieties (vertices) and blocks (arrays, edges). For microarrays we have particular interest in even designs, so the files also give the two best even designs when even designs exist. Every design is represented by its concurrence matrix, a v by v matrix whose i,j off-diagonal entry is 1 if varieties i and j appear in the same block, 0 otherwise. The diagonal entries in the concurrence matrix give the total replication of each variety (i.e. the degrees of the vertices).

Files: