Prediction of 8-state protein secondary structures by a novel deep learning architecture

BMC Bioinformatics

Table 1 Training and test data used in our work

Label	Types	TR6614		TR5534		CB513		CASP10		CASP11		CASP12
		Count	%	Count	%	Count	%	Count	%	Count	%	Count	%
H	α-helix	517653	0.352	405560	0.345	26143	0.309	6544	0.297	6330	0.309	3550	0.337
B	β-bridge	15321	0.010	12096	0.010	1180	0.014	227	0.010	221	0.011	113	0.011
E	β-strand	321156	0.218	255887	0.218	17994	0.212	5225	0.237	5089	0.248	2223	0.211
G	3₁₀helix	55994	0.038	46019	0.039	3132	0.037	797	0.036	716	0.035	320	0.030
I	π-helix	281	0	209	0	30	0	5	0	0	0	0	0
T	Turn	160753	0.109	132980	0.113	10008	0.118	2811	0.128	2299	0.112	1164	0.111
S	Bend	118800	0.081	97298	0.083	8310	0.098	1780	0.081	1751	0.085	955	0.091
L	Coil	282584	0.192	225493	0.192	17904	0.211	4652	0.211	4092	0.200	2201	0.209
All		1472542		1175542		84701		22041		20498		10526

ISSN: 1471-2105