Skip to main content

Table 2 Gene selection results of the simulation

From: A novel meta-analysis based on data augmentation and elastic data shared lasso regularization for gene expression

S

\(\rho\)

Lasso (%)

M-Lasso (%)

EN (%)

M-EN (%)

HLR (%)

M-HLR (%)

SGL (%)

Meta-Lasso (%)

DSL (%)

DA-DSL-\({\mathrm{L}}_{2}\) (%)

Youden’s index

1

0.3

60.52

92.63

83.31

95.88

90.18

91.31

94.58

93.27

94.78

96.79

 

0.6

46.08

87.90

47.51

88.58

42.34

78.99

86.74

85.71

86.71

92.32

 

0.9

41.02

66.80

49.60

83.26

44.34

86.34

73.84

74.52

72.65

91.29

2

0.3

43.56

80.44

45.43

83.02

48.95

84.72

83.51

84.41

83.21

85.89

 

0.6

42.08

76.38

42.53

80.82

48.25

82.10

82.16

80.38

78.18

84.15

 

0.9

41.98

66.31

41.88

66.41

40.75

67.70

63.43

65.91

65.09

69.78

3

0.3

43.37

72.97

43.27

73.13

42.83

69.66

76.80

76.52

78.74

84.42

 

0.6

49.46

63.47

56.72

81.46

44.80

80.11

73.59

71.15

72.12

82.37

 

0.9

38.66

57.47

42.25

75.52

39.08

75.53

63.72

62.60

61.57

80.08

4

0.3

46.50

50.39

50.24

51.63

47.31

50.42

59.81

57.50

56.23

68.74

 

0.6

43.53

70.93

43.48

80.29

41.03

79.17

76.26

77.00

79.76

86.17

 

0.9

43.58

56.48

45.82

73.03

43.73

75.92

65.62

70.76

72.93

79.78

  1. *S denotes the Scenario, and \(\rho\) is the correlation control variable of data. In bold – the best performance amongst all the methods