Table 2.

Comparing the performance of SPOT-1D-LM with single-sequence-based methods (SPIDER3-Single, ProteinUnet, and SPOT-1D-Single) and sequence-profile-based methods (SPOT-1D and NetSurfP-2.0) in the prediction of secondary structure in three (SS3) and (SS8) states, solvent accessibility (ASA), half-sphere-exposure-up (HSE-u), HSE-down (HSE-d), contact number (CN), backbone angles($\psi$, $\varphi$, $\theta$ and $\tau$) for TEST2018. Performance measures are accuracy for SS3 and SS8, correlation coefficient for ASA, HSE-u, HSE-d, and CN, and mean absolute errors for the angles.

Model	SS3	SS8	ASA	HSE-u	HSE-d	CN	$\psi$	$\varphi$	$\theta$	$\tau$
SPIDER3-Single	72.57	59.81	0.647	0.523	0.487	0.547	43.05	23.78	11.07	45.38
ProteinUnet	72.57	60.30	0.620	0.537	0.510	0.545	42.93	23.42	10.28	44.94
SPOT-1D-Single	74.28	72.17	0.665	0.573	0.563	0.585	40.58	22.16	9.35	42.32
NetSurfP-2.0(profile)	85.35	73.48	0.783	–	–	–	26.63	17.90	–	–
SPOT-1D (profile)	86.18	75.41	0.787	0.732	0.737	0.777	24.87	16.88	6.91	25.94
SPOT-1D-LM (This work)	86.74	76.47	0.814	0.759	0.761	0.690	23.74	15.99	6.46	24.60