Skip to main content

Table 4 Classification results on the validation and test sets of disease categories, averaged on all tasks within each category

From: MolData, a molecular benchmark for disease and target based machine learning

Benchmark

Validation Set

Test Set

Accuracy (%)

Recall (%)

Precision (%)

ROC AUC

Accuracy (%)

Recall (%)

Precision (%)

ROC AUC

All Tasks

64.7

76

3.93

0.7803

63.96

75.69

3.98

0.774

Cancer

73.61

68.76

3.68

0.7809

72.96

68.44

3.76

0.7765

Nervous System

73.34

65.1

2.39

0.7573

73.01

64.92

2.49

0.7556

Immune System

79.7

61.34

3.41

0.777

79.49

61.01

3.5

0.7739

Cardiovascular

80.06

56.84

2.98

0.7498

80.06

56.39

3.13

0.7457

Toxicity

86.9

33.41

24.46

0.7445

86.51

34.27

27.54

0.7309

Obesity

86.01

54.1

5.42

0.7925

85.37

51.5

5.51

0.7704

Virus

77.73

62.08

2.62

0.7625

77.9

61.91

2.84

0.7643

Diabetes

86.69

51.27

5.8

0.7845

85.88

51.33

5.99

0.7795

Metabolic Disorders

83.14

53.04

6.89

0.7619

82.71

54.85

7.06

0.7619

Bacteria

83.1

60.82

4.63

0.7916

82.26

64.49

4.69

0.8089

Parasite

91.51

46.65

11.31

0.8292

91.37

44.63

11.17

0.8243

Epigenetics-Genetics

88.46

45.27

6.36

0.7804

88.32

40.98

5.65

0.7251

Pulmonary

76.82

56.7

2.34

0.7293

76.06

54.79

2.5

0.7168

Infection

92.17

31.58

12.53

0.801

92.01

29.87

11.41

0.7871

Aging

94.83

23.59

1.86

0.7205

94.28

29.36

2.38

0.7402

Fungal

92.36

35.22

3.5

0.75

92.77

33.93

3.61

0.7335