Ranking non-synonymous single nucleotide polymorphisms based on disease concepts

Table 1 Performance of computational prediction algorithms when discriminating between disease-specific variants and other disease-causing/neutral variants

Algorithm	tp	fp	tn	fn	Accuracy	Precision	Specificity	Sensitivity	NPV	MCC	AUC
Musculoskeletal
SIFT	4,730	37,701	23,323	944	0.61	0.57	0.38	0.83	0.70	0.24	0.64
PolyPhen-2	5,278	44,047	34,859	714	0.66	0.61	0.44	0.88	0.79	0.36	0.71
FATHMM	5,902	51,596	29,202	201	0.66	0.60	0.36	0.97	0.92	0.41	0.73
Disease-Specific	4,120	3,123	77,675	1,983	0.82	0.95	0.96	0.68	0.75	0.66	0.93
Disease-Specific (20-fold)	-	-	-	-	0.80	0.92	0.94	0.66	0.74	0.63	-
Developmental
SIFT	845	41,586	23,983	284	0.56	0.54	0.37	0.75	0.59	0.12	0.56
PolyPhen-2	920	48,405	35,337	236	0.61	0.58	0.42	0.80	0.67	0.23	0.63
FATHMM	1,006	52,429	33,278	188	0.62	0.58	0.39	0.84	0.71	0.26	0.59
Disease-Specific	621	710	84,997	573	0.76	0.98	0.99	0.52	0.67	0.58	0.90
Disease-Specific (20-fold)	-	-	-	-	0.74	0.97	0.99	0.49	0.66	0.55	-
Endocrine
SIFT	3,084	39,347	23,443	824	0.58	0.56	0.37	0.79	0.64	0.18	0.60
PolyPhen-2	2,890	46,435	35,031	542	0.64	0.60	0.43	0.84	0.73	0.30	0.67
FATHMM	3,597	49,466	33,522	316	0.66	0.61	0.40	0.92	0.83	0.38	0.71
Disease-Specific	2,392	1,015	81,973	1,521	0.80	0.98	0.99	0.61	0.72	0.65	0.94
Disease-Specific (20-fold)	-	-	-	-	0.79	0.97	0.98	0.60	0.71	0.63	-
Metabolic
SIFT	10,731	31,700	21,913	2,354	0.61	0.58	0.41	0.82	0.69	0.25	0.64
PolyPhen-2	11,337	37,988	33,788	1,785	0.67	0.62	0.47	0.86	0.78	0.36	0.72
FATHMM	13,068	39,914	33,271	648	0.70	0.64	0.45	0.95	0.91	0.47	0.80
Disease-Specific	10,767	3,209	69,976	2,949	0.87	0.95	0.96	0.78	0.82	0.75	0.95
Disease-Specific (20-fold)	-	-	-	-	0.86	0.94	0.95	0.77	0.81	0.74	-

ISSN: 1479-7364