Skip to main content
Figure 2 | Genome Medicine

Figure 2

From: Sample-level enrichment analysis unravels shared stress phenotypes among multiple cancer types

Figure 2

Enrichment with the CIN signature reflects the response status to oncogene-induced stress and has prognostic power in breast cancer. (a) Heat map of tumor samples as columns and genes or modules as rows. Enrichment with gene modules is shown with colors from blue (downregulation) to red (upregulation) while gray values indicate no significant deviation from the expected median value. Samples are ordered according to the z-score of the chromosomal instability (CIN) signature. (Note that a sample might be insignificant (shown with gray) even if it has a z-score greater than the adjacent non-gray sample. This is because multiple testing correction is done for each sample independently.) Expression levels of genes are shown in colors from purple (low expression) to yellow (high expression). The samples with upregulation of the CIN signature include those that upregulate CDKN2A with concomitant down-regulation of RB1. Deregulation of Rb signaling in these samples is reflected in their over-expression of the Rb-E2F target genes CCNE1 and E2F3. These samples show a transcriptional program opposite to what is observed in senescent cells, indicating that they activated transcriptional programs indicative of senescence bypass. Shown is a breast cancer dataset by Ivshina et al. [21] (289 samples). (b) Kaplan-Meier curves for breast cancer patients. The red curve is for the samples with high expression of CKDN2A and MKI67 (lower panel) and positive enrichment of the CIN gene signature (upper panel). The black curves are for the rest of the samples in the dataset. It was shown that the concomitant expression of CDKN2A and MKI67 is related to impairment of the Rb pathway, hence to subsequent tumor events in ductal carcinoma in situ of the breast. The CIN gene signature has stronger prognostic power than the two-gene signature.

Back to article page