Skip to main content
. 2019 Jul 5;35(14):i389–i397. doi: 10.1093/bioinformatics/btz332

Fig. 4.

Fig. 4.

Comparison of evolutionary predictability among different cancer types. The vertical axis shows the CBN-based predictability ϕ, computed based on Equation (12) for each given cancer type. The error bars indicate the standard deviation of ϕ calculated from 100 bootstrap samples of equal size as the original genotype data. The genotypes for each cancer type are defined based on the mutational data of the corresponding n=10 most frequently mutated driver genes from TCGA (blue bars) or MSK-IMPACT (red bars). The cancer types are arranged from left to right in ascending order of their ϕ quantified based on TCGA data