. Author manuscript; available in PMC: 2010 Mar 1.

Published in final edited form as: Proteins. 2009 Mar;74(4):847–856. doi: 10.1002/prot.22193

Table 4.

The ten-fold-cross-validated accuracy for predicting the ϕ and ψ angles, and RSA. The comparison between prediction accuracies of this study and best reported accuracies

	ψ		ϕ		RSA
Criterion	Real-SPINE^a	This work^b	Real-SPINE^a	This work^b	Real-SPINE^c	This work^b
Q₁₀^d	46.6%	49.7%	45.0%	56.1%	34.7%	40.2%
Q_10%^e	64.7%	67.5%	80.1%	82.1%	-	58.2%
PCC^f	0.745	0.743	0.707	0.656	0.738	0.739
MAE^g	38.2°	36.4°	24.8°	22.1°	0.142	0.111

Real-SPINE 2.0²⁸ for the dataset of 2640 proteins.

Experiment IV for the dataset of 2640 proteins.

Real-SPINE²¹ for the dataset of 2640 proteins.

Q₁₀: Fraction of residues with correctly predicted states. Angles are divided into 10 states with 36° per bin. Residue solvent accessibility (RSA) is divided into 10 states with 0.1 per bin.

Q_10%: Fraction of residues whose angles are predicted within 36° (or 10% for RSA) from the true value.

Pearson’s correlation coefficient between predicted and actual values.

Mean-absolute error between predicted and actual values. Degrees are used for the ϕ and ψ angles and [0, 1] normalization for the RSA.