Skip to main content

Table 5 The designed molecules and their predicted dual-target activities

From: Perturbation-theory machine learning for mood disorders: virtual design of dual inhibitors of NET and SERT proteins

PREDICTIONSa, b,c, d

ID

cj

PredIAi(cj)

ProbAct(%)

ID

cj

PredIAi(cj)

ProbAct(%)

AMD-01

c01

1

66.32

AMD-03

c01

1

62.06

AMD-01

c02

-1

45.16

AMD-03

c02

1

62.87

AMD-01

c03

1

81.34

AMD-03

c03

1

81.98

AMD-01

c04

-1

30.67

AMD-03

c04

-1

35.10

AMD-01

c05

1

70.80

AMD-03

c05

1

50.68

AMD-01

c06

-1

47.31

AMD-03

c06

1

56.54

AMD-02

c01

1

57.73

AMD-04

c01

1

69.69

AMD-02

c02

1

75.61

AMD-04

c02

1

84.38

AMD-02

c03

1

79.94

AMD-04

c03

1

61.36

AMD-02

c04

-1

39.77

AMD-04

c04

1

65.81

AMD-02

c05

-1

40.66

AMD-04

c05

-1

35.44

AMD-02

c06

1

63.58

AMD-04

c06

1

55.30

  1. aID – Identifier for each designed molecule. bcj – The experimental conditions, which have been reported in the same order as in Table 3. cPredIAi(cj) – Predicted categorical value of activity against a specific mood-related protein (tp) and by considering a defined assay protocol (ai). dProbAct(%) – Probability (expressed in percentage) predicted by the PTML-MLP model for a molecule to be active