Skip to main content
Figure 7 | BMC Bioinformatics

Figure 7

From: Spatial normalization of array-CGH data

Figure 7

The proposed method (seg+2dLoess) compares favorably to all other normalization methods – breast cancer data set. We compared the proposed method (seg+2dLoess) to ten methods for two quality criteria: sigma and dyn. Each color corresponds to the comparison of seg+2dLoess with a different method. The proposed method is taken as a reference (red point 1 at (0, 0)). For each method i, the cross indicates the mean relative performance (see methods section) of the data set for dyn (x axis) and in sigma (y axis), and the lines give the corresponding 95% quantile of relative performance. For sigma (dyn, respectively), the methods with a 95% quantile below (left to, respectively) the horizontal (vertical, respectively) dashed black line are significantly outperformed by our proposed method. Here seg+2dLoess significantly outperforms all methods for dyn and sigma.

Back to article page