Fig. 2.

The eight eigengenes that are differentially expressed in both datasets. We computed an eigengene for each module as a weighted average of the expression of all genes in that module. The intensity of the colors in each heatmap corresponds to the normalized average expression. a Each column corresponds to an eigengene. Each row shows the expression of a case from the MILE microarray dataset. AML cases were first clustered and then plotted together with MDS cases for comparison. Three (five) eigengenes have clearly higher expression in MDS (AML) cases. Table 2 reports the adjusted p-values, shown in the green strips at the top. b The corresponding expression values are similar in the BCCA RNA-Seq dataset, which indicates that these biological signatures are robust and independent from the profiling platform