Skip to main content

Table 8 Precision of predictions as a function of the number of associated rules

From: A comprehensive study of small non-frameshift insertions/deletions in proteins and prediction of their phenotypic effects by a machine learning method (KD4i)

a) → Whole data set

N° of rules

N° of deleterious NFS-Indels covered

N° of neutral NFS-Indels covered

Average precision

1

27

5

84.38

2

26

7

78.79

3

25

1

96.15

4

16

1

94.12

5

14

3

82.35

6

13

1

92.86

7

220

2

99.01

b) → Test set

N° of rules

N° of deleterious NFS-Indels covered

N° of neutral NFS-Indels covered

Average precision

1

8

5

61.54

2

6

7

46.15

3

6

1

85.71

4

8

1

88.89

5

4

3

57.14

6

3

1

75.00