| |
Validation sets
|
|---|
|
Methods
|
C (9 genes)
|
B (18 genes)
|
M (25 genes)
|
| |
AUC
|
( /η
max
)
|
AUC
|
(/η
max
)
|
AUC
|
(/η
max
)
|
|
Statistical CGP (scoring functions)
|
|
sens
|
0.858
|
(2.0/5.4)
|
0.853
|
(1.9/4.3)
|
0.830
|
(1.8/3.8)
|
|
spec
|
0.396
|
(0.5/1.5)
|
0.427
|
(0.7/2.6)
|
0.506
|
(1.1/5.2)
|
|
ppv
|
0.420
|
(0.6/1.57)
|
0.504
|
(1.3/29.5)
|
0.590
|
(2.1/85.0)
|
|
npv
|
0.966
|
(3.6/30.1)
|
0.964
|
(3.5/21.7)
|
0.978
|
(3.2/17.3)
|
|
amss
|
0.985
|
(4.6/88.5)
|
0.980
|
(4.4/59)
|
0.970
|
(4.4/85.0)
|
|
hmss
|
0.986
|
(4.8/88.5)
|
0.980
|
(4.5/64)
|
0.969
|
(4.5/85.0)
|
|
OR
|
0.415
|
(0.5/1.57)
|
0.509
|
(1.3/29.5)
|
0.592
|
(2.1/85.0)
|
|
chisq
|
0.978
|
(4.2/59.0)
|
0.975
|
(3.9/34.7)
|
0.959
|
(3.7/28.3)
|
|
bchisq
|
0.978
|
(4.2/59.0)
|
0.975
|
(3.9/34.7)
|
0.960
|
(3.7/28.3)
|
|
F
|
0.932
|
(3.3/32.9)
|
0.915
|
(3.1/23.1)
|
0.881
|
(2.8/18.5)
|
|
Inductive CGP (machine learning algorithms)
|
|
NB
|
0.901
| |
0.879
| |
0.843
| |
|
LR
|
0.980
| |
0.905
| |
0.887
| |
|
ADTree
|
0.996
| |
0.944
| |
0.975
| |
|
IBk
|
0.948
| |
0.950
| |
0.974
| |
|
J48
|
0.885
| |
0.832
| |
0.752
| |
|
SMO/Poly
|
0.999
| |
0.948
| |
0.879
| |
|
SMO/RBF
|
0.998
| |
0.991
| |
0.909
| |
- This table lists the performance of statistical and inductive CGPs in prioritising peptidoglycan-related genes in Streptococcus agalactiae 2603 V/R. Abbreviations: sens: sensitivity; spec: specificity; ppv: positive predictive value; npv: negative predictive value; amss: arithmetic mean of sensitivity and specificity; hmss: harmonic mean of sensitivity and specificity; OR: odds ratio; chisq: chi-square; bchisq: signed chi-square; F: F-measure; NB: naïve Bayes classifier; LR: logistic regression; ADTree: alternating decision tree; IBk: k-nearest neighbour classifier; J48: J48 decision tree; SMO: support vector machine trained by sequential minimal optimisation algorithm; Poly: polynomial kernel; RBF: radial basis function kernel.
/η
max
)