Skip to main content
. 2018 Apr 26;8:6620. doi: 10.1038/s41598-018-24937-4

Figure 2.

Figure 2

Samples from different cancer types and their random partition into training and test sets. Each sample (represented using a vector of 1239 TFs expressions) is mapped onto a 2-dimensional space using multidimensional scaling. Samples of various cancer types of PanCan12 dataset are shown in different colors. Circles represent test samples (i.e., test conditions) and crosses represent training samples.