Table 6 Comparative table of statistical metrics of each model for the training set (Table S2).

From: Identification of flux checkpoints in a metabolic pathway through white-box, grey-box and black-box modeling approaches

 

R2

RMSE

MAE

AIC

PGAM

ENO

PPDK

R2

RMSE

R2

RMSE

R2

RMSE

Model 0

0.85

4.33

3.17

584.74

0.98

2.42

1

2.41

0.71

6.75

Model 1

0.88

3.39

2.48

494.39

0.98

2.02

0.98

1.78

0.8

5.27

Model 2a

0.71

4.23

2.78

99.5

0.94

2.19

0.78

4.02

0.41

5.71

Model 3a

1

0.52

0.37

124.21

1

0.62

1

0.62

1

0.22

Model 4

0.94

2.43

2.1

396.72

0.98

2.96

0.97

2.3

0.94

1.94

Model 5

0.95

2.06

1.7

336.05

0.98

2.59

0.98

1.89

0.96

1.6

Model 6

0.98

1.71

1.47

244.71

0.98

2.02

0.98

1.78

0.99

1.22

  1. RMSE and MAE are in nmol·min−1.
  2. aFor these models, (cv)RMSE and (cv)R2 are calculated.
Back to article page