Figure 4
From: A distribution-free convolution model for background correction of oligonucleotide microarray data
ROC Curves for HG-U95Av2 Latin Square Spike-In Experiment for d = 1 (a) and d = 2 (b). For other values of d, the differences among methods are very small. Curves for MAS 5.0 and PLIER (and dChip for plot b) do not appear on the graphs because their false positive and true positive rates are too small for the scale given. The scale goes from 0.4 to 0.9 in order to magnify differences among the methods.