Genes identified though screening are differentially expressed across breast cancer subtypes. (A) Expression levels of the 47 genes found to be required for mammary epithelial cell growth are displayed on a heat map representing 1,998 breast cancer patients within the METABRIC dataset. Gene expression was determined on the Illumina HumanHT-12 Expression BeadChip array (Illumina, San Diego, CA, USA). Cases are clustered according to ten biological primary breast cancer subgroups described in Additional file 2: Figure S8 (also see ). (B) Beanplots depicting target gene expression in the ten biological primary breast cancer subgroups for RIPK2, TUFT1, EFNA4 and TMEM9B. The individual observations are shown as small horizontal lines in a one-dimensional scatterplot with the estimated density of the distributions shown in colour and the average indicated by the long horizontal line. (C) Rank-ordered plot of hazard estimates and unadjusted 95% confidence intervals for high gene expression relative to low gene expression demonstrates differences in survival based upon the expression of the target genes. Gene targets are marked with identifiers for binarized expression variables showing adjusted P-values <0.05 (false discovery rate (FDR), 0.05) (denoted by ** or ++) and adjusted P-values <0.1 (FDR, 0.1) (denoted by * or +). Single asterisk indicates a significant finding after adjustment for multiple comparisons using the G-rho model test (ρ = 1). Single plus sign indicates a significant finding after adjustment for multiple comparisons using an omnibus test from a Cox model with biomarker and IntClust subtype and their interaction terms. K-W, Kruskal-Wallis test.