GPU-based detection of protein cavities using Gaussian surfaces

BMC Bioinformatics

Table 1 Performance of benchmarking detection methods for apo proteins in terms of: (d) distance (FN) false negatives to PDBsum ground-truth cavity centers; (TP) true positives; (FP) false positives; (TN) true negatives; (S _v) sensitivity; (S _c) specificity; (a) accuracy; (r _d) ratio of detected ground-truth cavities; and (C _u) cumulative number of undetected ground-truth cavities

	GaussianFinder	ConCavity	POCASA	SURFNET	PASS	GHECOM	Fpocket	3V	KVFinder
d∈[0.0,1.0]	7100	1512	2310	1737	4915	4562	5869	3751	4578
d∈]1.0,2.0]	188	107	176	1227	884	255	195	417	476
d∈]2.0,3.0]	227	127	226	520	117	380	207	291	305
d∈]3.0,4.0]	182	203	297	432	157	413	257	148	193
TP	7697	1949	3009	3014	6073	5610	6528	4607	5552
FP	1033	2103	863	902	325	581	878	1017	979
TN	2045	860	1697	1247	3086	2540	2827	1782	1889
FN	393	207	289	216	44	500	148	428	467
S _v	0.951	0.904	0.912	0.933	0.993	0.918	0.978	0.915	0.922
S _c	0.664	0.290	0.663	0.580	0.905	0.814	0.763	0.636	0.659
a	0.872	0.549	0.803	0.792	0.961	0.883	0.901	0.815	0.837
r _d	0.944	0.239	0.369	0.369	0.745	0.688	0.801	0.565	0.681
C _u	453	6201	5141	5136	2077	2540	1622	3543	2598

ISSN: 1471-2105