Fig. 6.

C1 and C2 form smaller sub-clusters of genes in the TCGA clinical dataset. Spearman’s autocorrelation amongst the genes in a C1 and b C2 over the 1026 breast tumour samples provided by the TCGA database. Indexing of the rows starts from top to bottom and for the columns starts from right to left, i.e. the first gene is represented by the top-most row and the right-most column. Colours in the heat map range from dark brown for +1.0 to dark blue for −1.0. Sub-clusters of genes based on this autocorrelation are labelled to the right of the heat maps and white lines are drawn over the heat maps to highlight their borders