Table 1 The comparative results of the different models: MLR, RSM, and MLRSM scenarios 1–3.

From: Understanding the hidden relations between pro- and anti-inflammatory cytokine genes in bovine oviduct epithelium using a multilayer response surface method

Gene

Statistics

MLR

RSM

MLRSM scenario 1

MLRSM scenario 2

MLRSM scenario 3

IL1B

RMSE

4.64E-3

4.27E-3

3.08E-3

1.23E-3

2.13E-3

MBE

46.44

19.42

12.11

0.84

4.40

d

0.42

0.59

0.85

0.98

0.95

EF

0.11

0.25

0.61

0.94

0.81

TNFA

RMSE

1.11E-2

1.07E-2

9.53E-3

6.84E-3

7.46E-3

MBE

2.70

2.61

2.63

1.04

1.05

d

0.28

0.31

0.74

0.88

0.82

EF

−0.05

0.03

0.23

0.34

0.53

TLR4

RMSE

3.41E-4

3.35E-4

2.96E-4

2.52E-4

2.73E-4

MBE

0.95

0.53

0.28

0.22

0.19

d

0.74

0.76

0.82

0.89

0.87

EF

0.35

0.38

0.51

0.65

0.58

IL10

RMSE

9.12E-5

8.73E-5

7.83E-5

3.69E-5

4.19E-5

MBE

533.95

517.29

17.00

4.50

6.95

d

0.30

0.35

0.56

0.95

0.93

EF

−0.13

−0.04

0.16

0.81

0.76

IL4

RMSE

3.73E-5

4.81E-5

3.06E-5

1.76E-5

2.87E-5

MBE

1.91

−0.06

0.15

0.09

0.36

d

0.34

0.41

0.65

0.92

0.71

EF

−0.09

−0.81

0.27

0.76

0.35

  1. Bold numbers are the best statistics obtained for each model as well as each gene. RMSE is the root-mean-square errors, MBE is the mean bias error, EF is the Nash-Sutcliffe efficiency, and d is Willmott’s index of agreement.
Back to article page