Figure 4

Relevance of selected gene expression probe sets to the top CCA features. Graphical representation of the cross-loadings, indicating the relevance of the most important gene expression probe set to the first two features extracted using regularized dual CCA, with regularization parameters τ _x = 0.9, τ _y = 0.3 (left panel), and PCA+CCA (right panel). The value on the horizontal axis indicates the cross-loading with the first feature, and the value on the vertical axis indicates the cross-loading with the second feature. The circle has radius 1, indicating perfect correlation between the expression of a probe set and a copy number feature. The lines show the cross-loadings for the probe sets with the highest relevance to each of the two features, and the genes corresponding to the ten probe sets with highest positive and negative correlation with each feature, respectively, are given. The overall highest correlation is found with regularized dual CCA (left panel), between a probe set corresponding to the PBX1 gene and the second feature.