Skip to main content

Table 4 Comparison of the results reported for the 25PDB benchmark (in percentage %)

From: Proposing a highly accurate protein structural class predictor using segmentation-based features

References

Method

All-α

All-β

α/β

α+ β

Overall

[45]

Logistic Regression

69.1

61.6

60.1

38.3

57.1

[53]

Specific Tri-peptides

60.6

60.7

67.9

44.3

58.6

[33]

LLSC-PRED

75.2

67.5

62.1

44.0

62.2

[33]

SVM

77.4

66.4

61.3

45.4

62.7

[38]

AAD-CGR

64.3

65.0

65.0

61.7

64.0

[7]

CWT-PCA-SVM

76.5

67.3

66.8

45.8

64.0

[54]

AATP

81.9

74.7

75.1

55.8

71.7

[16]

AADP-PSSM

83.3

78.1

76.3

54.4

72.9

[55]

SCPRED

92.6

80.1

74.0

71.0

79.7

[37]

SSA

92.6

83.7

80.5

65.9

81.5

[37]

PSSA

94.6

76.3

73.1

74.4

80.0

[24]

RKS-PPSC

92.8

83.3

80.8

70.1

82.9

[48]

SVM

92.6

81.3

81.5

76.0

82.9

[27]

MODAS

92.3

83.7

81.2

68.3

81.4

[26]

AAC-PSSM-AC

85.3

81.7

73.7

55.3

74.1

[22]

Physicochemical-based features

86.1

80.8

80.6

60.1

76.7

[5]

Structural-based features

95.0

85.6

81.5

73.2

83.9

[6]

Structural-based features

95.0

81.3

83.2

77.6

84.3

This Study

PSSM-S

93.5

90.3

92.1

81.4

89.6

This Study

SPINE-S

93.8

83.1

78.4

73.9

82.3

This Study

PSSM-SPINE-S

96.8

93.7

90.1

87.0

92.2