Skip to main content

Table 4 The AUCs values of different graph convolutional networks methods over 16 diseases for threshold = 5

From: A novel candidate disease gene prioritization method using deep graph convolutional networks and semi-supervised learning

Disease

Methods

GCNGP

GCAS

RGCN

PGCN

GCN-MF

Pancreatitis

80.76

73.38

76.18

78.01

78.57

Parkinson’s disease

75.28

71.25

70.90

74.59

73.16

Celiac disease

77.95

73.81

70.39

71.27

75.82

Atherosclerosis

80.25

76.66

73.11

75.05

75.98

Esophageal cancer

78.61

75.15

74.77

75.86

79.10

Crohn’s disease

69.90

65.49

63.26

65.07

67.44

Breast cancer

74.08

71.69

71.87

70.78

72.88

Alzheimer’s disease

76.39

73.94

70.85

69.97

74.29

Ulcerative colitis

67.83

65.81

62.19

62.92

64.66

Endometriosis

84.30

81.17

79.67

79.95

83.13

Cirrhosis

59.92

56.94

55.09

56.19

57.75

Myocardial infarction

71.48

70.18

67.75

66.54

69.49

Tuberculosis

78.75

75.97

71.54

70.72

76.91

Lymphoma

73.87

72.45

69.19

70.04

71.97

Rheumatoid arthritis

66.09

64.93

63.17

61.59

64.68

Asthma

63.94

60.74

59.13

59.91

61.39

Average

73.71

70.60

68.69

69.28

71.70