Figure 7

Hierarchical clustering analysis. Hierarchical clustering analysis using both DNA (a) and RNA (b) datasets. The bootstrapping procedure estimates how strongly the clusters are supported by data. Bootstrap values are estimated using multi-scale bootstrap resampling (shown in red numbers) and normal resampling (shown in green numbers), and are reported as percentages, indicating how often a cluster was observed in the bootstrapping [34]. P = primary tumor, M = metastasis, N-normal tissue. For the RNA clustering plot, P = primary tumor, M = metastasis, N = normal tissue. Note: for patient 231 two separate brain metastases were analyzed and both clustered together as shown.