Clustering analysis of translational efficiencies. (A) The TE values for each detected gene were used to cluster samples using Mcquitty linkage with a Manhattan distance matrix; values at branch points are approximately unbiased (AU) P values, calculated by multiscale bootstrap resampling using the pvclust R package. Patients 18, 24, 27, 28, 30, and 31 (type A) are denoted within the green rectangle; patients 25 and 29 (type B) are denoted within the red rectangle. (B) The TE values for each detected gene were used in a principal components analysis. The outlying patients (10 and 22) are illustrated as blue dots, while type A are green dots and type B are red dots.