Figure 2.

Breast cancer cell lines can be hierarchically clustered according to their expression of the MAPK-related gene predictors. Each row represents the relative transcript abundance (in log₂ space) for one gene; each column represents data from one cell line. Expression of genes in the top panel predicts sensitivity to MEK inhibitors, whereas expression of those in the bottom panel predicts resistance. Within each panel, genes are ordered by q value. In the top panel, the most significant predictors are at the top; in the bottom panel, the most significant predictors are at the bottom. In the tree, yellow end nodes denote basal cell lines and pink end nodes denote luminal cell lines.