Fig. 3From: MUFFINN: cancer gene discovery via network analysis of somatic mutation dataMUFFINN performs best by using mutation information of direct neighbors only. Performance assessment was conducted similarly to those of Fig. 2 for different network algorithms, including three diffusion algorithms on HumanNet: distributions of 18 AUC scores for cancer genes annotated by a CGC and b the 20/20 rule, and cumulative numbers of retrieved cancer genes annotated by c CGC and d the 20/20 rule within the top 100, 500, and 1000. MUFFINN shows higher predictive power for cancer genes by using mutation information of direct neighbors only than by using all genes of HumanNet with various network diffusion algorithmsBack to article page