Figure 2
miRNA and mRNA scores differ between basal and luminal breast cancers. (A,B) are plots related to miRNAs, while (C,D) are related to mRNAs. Plots in A and C are based on data from the present study and from GSE68256, respectively, data in B are from TCGA and data in C from35. We identified 18 up-regulated and 10 down-regulated miRNAs, which are significantly deregulated in p95 samples and not in HER2 samples (red points). While there is a tight relationship between mRNAs upregulated in p95HER2 and HER2 (slope = 1.46, p-value = 5.06e-142) (panel C), surprisingly we did not find the same relationship for miRNAs (slope = 0.0442, p-value = 0.375) (panel A). We used the significant miRNAs from the present study to calculate a miRNA score: miRNA Score = mean(up-regulated miRNAs) − mean(down-regulated miRNAs). This score was calculated for all samples in the TCGA cohort and we found a significantly (Wilcoxon) greater score for samples from basal compared to luminal tumor types (PAM50). There are 50 downregulated and 53 upregulated mRNAs in the mRNA score calculated from GSE68256 as: mRNA Score = mean(filtered up-regulated mRNAs) − mean(filtered down-regulated mRNAs). For both miRNA (B) and mRNA (D), scores were higher in basal than in luminal samples indicating that both mRNAs and miRNAs deregulated in the MCF-7 cells (luminal) are involved in the basal signature for breast cancer.
