Table 2.

Comparative Performance on cross-validation between BRCA1/2 classification models and other available approaches.

	Methods	ACC	MCC	Sens	Spec	ROC-AUC
BRCA1	BRCA1-Ring domain41	–	–	95%	100%	0.99
	MLR42	0.82	0.65	69%	93%	–
	MLR-CAGI42	0.75	0.50	61%	87%	–
	SIFT12	0.68	0.50	92%	55%	0.73
	PolyPhen-215	0.64	0.40	55%	69%	0.62
	Align-GVGD19,20	0.48	0.18	48%	26%	0.58
	REVEL13	0.68	0.48	68%	55%	0.74
	CADD40	0.74	0.58	74%	61%	0.80
	BRCA1	0.96	0.96	93%	98%	0.98
	BRCA1-ENIGMA	0.99	0.96	99%	99%	0.98
BRCA1-combined	SGE38	–	–	96.7%	98.2%	0.98
	BRCA-ML11	–	0.66	89.5%	91.5%	–
	BRCA1-Ring domain41	–	–	97%	83%	0.96
	NN42	0.87	0.75	92%	85%	–
	NN-CAGI42	0.76	0.55	86%	71%	–
	SIFT12	0.68	0.50	92%	55%	0.74
	PolyPhen-215	0.64	0.40	55%	69%	0.62
	Align-GVGD19,20	0.40	0.18	38%	23%	0.54
	REVEL13	0.68	0.48	68%	55%	0.74
	CADD40	0.74	0.58	74%	60%	0.80
	Our model	0.97	0.96	94%	98%	0.99
BRCA2	MLR42	0.78	0.57	82%	74%	–
	MLR-CAGI42	0.86	0.71	86%	85%	–
	SIFT10	0.64	0.45	96%	58%	0.77
	PolyPhen-213	0.61	0.43	89%	56%	0.72
	Align-GVGD19,20	0.31	0.18	31%	24%	0.47
	REVEL13	0.85	0.48	85%	89%	0.77
	CADD40	0.76	0.58	76%	71%	0.86
	BRCA2	0.98	0.98	96%	98%	0.97
	BRCA2-ENIGMA	0.99	0.95	99%	99%	0.99
BRCA2-combined	BRCA-ML9	–	0.73	97.7%	85.1%	–
	NN42	0.84	0.59	75%	86%	–
	NN-CAGI42	0.76	0.47	75%	77%	–
	SIFT10	0.72	0.45	93%	68%	0.80
	PolyPhen-213	0.68	0.40	88%	64%	0.76
	Align-GVGD19,20	0.30	0.18	30%	22%	0.49
	REVEL13	0.87	0.54	87%	91%	0.79
	CADD40	0.80	0.54	80%	77%	0.85
	Our model	0.91	0.89	82%	93%	0.92
BRCA1/2 general model	BRCA-ML9	–	–	74%	98%	–
	SIFT10	0.70	0.48	92%	63%	0.78
	PolyPhen-213	0.66	0.40	66%	66%	0.66
	Align-GVGD19,20	0.38	0.06	38%	24%	0.53
	REVEL13	0.75	0.50	75%	72%	0.79
	CADD40	0.78	0.60	78%	72%	0.84
	Our model	0.96	0.91	96%	98%	0.96