Expression profiles of pediatric AMLs. (A) Hierarchical clustering of 130 diagnostic pediatric AML samples (columns) versus 250 class discriminating genes (rows). The genes used in this analysis are the top 50–ranked genes per group as selected by SAM. For genes that had more than one probe set selected as a class discriminator, the highest-ranked probe set was used for this figure. Probe set signal values were normalized to the mean across the entire data set and the relative value for each case is represented by a color, with red representing high expression and green representing low expression (scale shown in the lower right). The genetic subtype of each case is indicated by colored bars across the top and bottom of the panel. (B) Similarity plot of 130 pediatric AML diagnostic samples using the top 50–ranked genes (1 probe set per gene) for each subgroup as selected by SAM. Similarities are plotted using a scale that is based on Pearson correlation coefficients calculated for pairwise comparisons using the expression data. The degree of similarity between cases is displayed using the blue color scale at the bottom of the figure. Genetic groups are indicated by the color bars along the top and side of the similarity plot and are arranged identically as in panel A.