Figure 1.

Hierarchical clustering and data scatter. (A) Heat map of the hierarchical clustering and SOM analyses of all data. (B) Heat map of the clustering and SOM analyses of the erythroid and megakaryoblastic over-expressed genes. Each column represents a leukaemia (T-cell: T1, T2, T3; B-cell: B1, B2, B3; Myeloid: M; Erythroid: E1, E2, E3; Megakaryoblastic: Mk1, Mk2, Mk3) and each row is assigned to a transcript. A red colour means a positive deviation of 0.595 and above from the mean (over-expression) and a green colour means a negative deviation of -0.595 and below from the mean (under-expression). The black colour corresponds to values comprised between 0.595 and -0.595. The result of the clustering on the arrays is shown as a dendrogram on the top of the figure. The nodes of the tree indicate the degrees of similarity between the leukaemias.