Table 1. Comparison of different methods to distinguish neutral from non-neutral mutations.

From combined experimental dataset. Mutations were observed in corresponding cancer cohorts. See S6 and S7 Tables for results on rare and all mutations. Maximum Matthew’s correlation is reported for each predictor which are ranked with respect to the maximum Matthew’s correlation coefficient. B-Score for each cohort is calculated with the respective cohort size: COSMIC v85 cohort 12,013; MSK-Impact 9,228. For CHASM the background model yielding best performance was chosen.

Measure	AUC-ROC	AUC-PR	Matthew’s correlation	Sensitivity at 10% FPR	Cohort
CHASMplus	0.91	0.95	0.64	0.74	COSMIC v85
MutaGene	0.88	0.94	0.61	0.65
CanDrA	0.82	0.88	0.58	0.49
FatHMM	0.84	0.90	0.56	0.53
REVEL	0.84	0.91	0.51	0.60
Frequency	0.79	0.90	0.49	0.60
CHASM	0.75	0.83	0.48	0.20
VEST	0.78	0.87	0.45	0.40
CHASMplus	0.90	0.94	0.66	0.75	MSK-IMPACT
FatHMM	0.86	0.90	0.66	0.53
MutaGene	0.89	0.94	0.65	0.69
CanDrA	0.83	0.87	0.62	0.42
REVEL	0.87	0.92	0.58	0.62
Frequency	0.80	0.90	0.51	0.62
CHASM	0.79	0.86	0.47	0.38
VEST	0.79	0.87	0.45	0.41