Skip to main content

Table 2 Overall accuracies obtained with different combinations of feature subsets.

From: Prediction of protein structural classes for low-homology sequences based on predicted secondary structure

Dataset

R

K

S

R+ K

R+ S

K+ S

R+ K+ S

25PDB

81.4

76.0

72.7

81.2

82.3

78.2

82.9

640

80.5

77.2

73.9

81.1

82.5

78.9

83.1

1189

79.6

75.8

73.0

80.3

81.0

79.7

81.3

  1. See text for the notations of R, K and S.