Skip to main content

Table 4 The prediction performance with different window sizes.

From: Prediction of RNA-binding amino acids from protein and RNA sequences

Window size

Sn (%)

Sp (%)

Pr (%)

Ac (%)

NP (%)

Fm (%)

CC

1

47.3

73.6

16.1

71.1

60.4

57.6

0.14

3

66.5

84.8

30.9

82.7

74.1

72.6

0.36

5

64.9

84.5

30.9

82.6

74.7

73.4

0.36

7

65.6

84.9

31.7

83.0

75.2

74.0

0.37

9

66.8

84.9

32.2

83.2

75.9

74.8

0.38

11

66.8

85.4

32.8

83.6

76.1

74.9

0.39

13

67.1

85.8

33.7

84.0

76.5

75.3

0.40

15

68.4

85.9

34.1

84.2

77.1

76.1

0.41

  1. The non-redundant dataset with various window sizes were constructed by applying the feature vector-based redundancy reduction method to the PRI3149 dataset. All 9 features were encoded in the feature vector. The prediction performance across varying window sizes was evaluated by 5-fold cross validation. Sn: sensitivity. Sp: specificity. Pr: precision rate. Ac: accuracy. NP: net prediction. Fm: F-measure. CC: correlation coefficient.