Skip to main content

Table 4 Scenario II simulation results of binary group variable in the univariable robust regression model using pseudo-value approach with 1000 replicates

From: A pseudo-value regression approach for differential network analysis of co-expression data

p

n

Precision

Recall

F1

PRANA

dnapath

DINGO

PRANA

dnapath

DINGO

PRANA

dnapath

DINGO

20

40

0.90

0.97

0.89

0.59

0.38

0.63

0.69

0.53

0.72

100

0.91

0.97

0.88

0.72

0.66

0.75

0.79

0.77

0.80

200

0.91

0.96

0.89

0.83

0.83

0.73

0.86

0.88

0.79

500

0.90

0.93

–

0.93

0.94

–

0.91

0.93

–

1000

0.89

0.91

–

0.95

0.97

–

0.92

0.94

–

50

40

0.98

1.00

0.99

0.56

0.28

0.67

0.71

0.43

0.80

100

0.98

1.00

0.98

0.65

0.45

0.71

0.78

0.61

0.82

200

0.99

1.00

0.98

0.75

0.68

0.80

0.85

0.80

0.88

500

0.99

0.99

–

0.87

0.91

–

0.93

0.95

–

1000

0.99

0.99

–

0.94

0.97

–

0.96

0.98

–

100

40

0.99

1.00

0.55

0.55

0.21

0.77

0.70

0.34

0.63

100

0.98

1.00

0.55

0.63

0.27

0.75

0.76

0.42

0.63

200

0.99

1.00

0.55

0.68

0.46

0.77

0.80

0.62

0.63

500

0.99

1.00

–

0.77

0.82

–

0.86

0.90

–

1000

0.99

1.00

–

0.86

0.92

–ara>

0.92

0.96

–

  1. The network structure depends on age covariate. Random network is generated at each simulation replicate. Sample size \(n = (500, 1000)\) or gene size \(p = 100\) were not included for DINGO due to heavy computational time. The best results are highlighted in boldface