Figure 2

Clustering of arrays visualised by two different methods. A. GE-biplot of the 9098 probesets (red asterisks) with the first 50 (1 = most significant), as ranked by ANOVA, indicated. The position of each array is shown as a label joined by a dotted line to the origin. B. A heat map (white: up-regulated; red: down-regulated) using hierarchical clustering with complete linkage for the columns only. The rank order of the p-values, gene names and Affymetrix IDs are included on the right and correspond to the numbering in A.