Skip to main content
. 2015 Feb 21;16(1):40. doi: 10.1186/s13059-015-0599-z

Figure 7.

Figure 7

Global scenario of subphenotypes associated with different tumor pathophenotypes and clusters of prognosis. (A,B) Networks of subphenotypes associated with latency (A) and number of metastases (B). Red edges indicate a negative correlation, green edges a positive correlation. All simple correlations with P < 0.05 and r > 0.3 were included (Cytoscape Software). Other examples are shown in Figure S7B in Additional file 2. (C) This figure summarizes the distribution of tumor pathophenotypes and different levels of subphenotypes through the clusters of prognosis. The values of the data integrated here can be found in Tables S3, S5, S10 and S20 in Additional file 1. In the case of tumor pathophenotypes, signaling pathways and metabolites, the median is represented as a percentage of the highest value in each row. For genetic markers the percentage of mice that are heterozygotic in each cluster is shown. Non-applicable traits are represented in black.