Figure 7

Support Trees. Hierarchical clustering using MeV software with 100 bootstrap iterations on samples and experiments was used to generate a dendrogram. For display purposes the most stable Upper Node (left) and Lower Node (right) are separated for display purposes. Branches in the dendrogram are color-coded according to reliability (as a percentage of all bootstrap iterations in which cluster was identified: Black- 100%, Gray 90–100%, Blue- 80–90%, Green- 70–80%, Yellow- 60–70%, Orange- 50–60%, Fuscia- 0–50%, Red- 0% and Pink- unrecovered), and shaded triangular areas are used to consolidate subjects/genes for display. Beneath each section of the dendrogram is a graph depicting mean standardized intensities for highlighted groups of subjects that were reliably clustered with one another. Below the graphs are representative Gene Ontology over-expressed Biological Processes for Upper and Lower nodes. Gene expression intensities are expressed in standard deviations from the mean for each gene (see scale bar).