Skip to main content

Table 1 Performance for each feature set by 5-fold cross-validation.

From: Predicting disease-associated substitution of a single amino acid by analyzing residue interactions

 

Sensitivity (%)

Specificity (%)

ACC (%)

MCC

All feature set

(200 a,2b)

89.8

72.7

83.0

0.64

37-feature set

(200,3)

91.0

72.3

83.6

0.65

f-set 1c

(100,4)

79.7

65.4

74.1

0.45

f-set 2d

(300,3)

81.2

63.6

74.3

0.45

f-set 3e

(200,1)

85.4

67.0

78.1

0.54

  1. Optimized parameters for random forest are listed in parentheses. Detailed description of each feature set is given in the Results section.
  2. a the optimal value of ntree (the number of trees to be grown).
  3. b the optimal value of mtry (the number of variables selected to determine the decision at a node of the tree).
  4. c the conservation feature set.
  5. d the network feature set.
  6. e the environmental feature set.