Figure 3
From: Intrinsic bias in breast cancer gene expression data sets
ER-α expression account for more variance in the expression of proliferation markers in the NKI2 and TRANSBIG data sets. A. The graph shows the percentage decrease in the sum of the variances of all genes within a data set after globally adjusting that data set for either ER-α expression (grey bars) or a marker of proliferation (black bars). B. The graph shows the percentage of the variance associated with adjustment for proliferation that is accounted for by adjustment for ER-α expression. This percentage reflects the extent to which genes whose expression follows proliferation-associated genes also fluctuate in response to changes in ER-α expression.