Table 4.

Accuracy of MLb-LDLr, PolyPhen-2, SIFT, SFIP-MutID, CADD, and MutationTaster on 13 Characterized Novel LDLr Variants

Software	Predicted Pathogenic (n = 11)	Predicted Pathogenic, %	Predicted Benign (n = 2)	Predicted Benign, %	General Accuracy, %
MLb-LDLr	8	72	1	50	69
PolyPhen-2	6	54	0	0	46
SIFT	2	18	2	100	30
SFIP-MutID	10	91	0	0	77
CADD	6	54	0	0	46
MutationTaster	8	72	0	0	61

Values are n unless otherwise indicated.

MLb-LDLr = machine-learning–based low-density lipoprotein receptor software; other abbreviations as in Tables 1 and 2.