Skip to main content

Table 2 Comparison of AUC for GEBN and logistic regression

From: Identification of genetic interaction networks via an evolutionary algorithm evolved Bayesian network

 

Functional SNPs in Case data

Functional SNPs in Control data

Weight (W)

MAX

Regression

GENN

GEBN

100

100

500

100

500

100

500

Main effect

SNP A

SNP A

0.1

55

52

51

54

54

53

52

0.5

71

71

64

70

71

71

67

0.9

88

88

72

88

88

88

87

SNP A, B, C, D

SNP A, B, C, D

0.1

61

57

53

54

54

58

54

0.5

90

89

72

73

73

89

87

0.9

99

96

87

98

98

99

99

Interaction effect

SNP A * SNP B

None

0.1

53

50

50

50

50

50

50

0.5

67

50

49

56

53

65

60

0.9

89

50

50

60

59

80

77

SNP A * SNP B

SNP W * SNP X

0.1

56

50

50

50

50

52

51

SNP C * SNP D

SNP Y * SNP Z

0.5

82

50

50

57

55

81

67

0.9

97

49

50

62

60

97

89