Skip to main content

Table 6 The identified fragments by the information gain (IG) method couple with frequency analysis for BCRP inhibitors

From: ADMET evaluation in drug discovery. 20. Prediction of breast cancer resistance protein inhibition through machine learning

Fragment

Name

Count1a

Count2b

F1c

F2d

IG

Quinazoline

175

14

2.3604

0.1219

0.0665

N,2-diphenylquinazolin-4-amine

92

0

2.5492

0.0000

0.0456

Pyrazolo[1,5-a]pyrimidine

85

8

2.3299

0.1415

0.0299

4H-chromene

104

23

2.0875

0.2980

0.0263

2-phenyl-4H-chromene

55

3

2.4173

0.0851

0.0216

2H-tetrazole

65

12

2.1519

0.2564

0.0177

Piperazine

90

30

1.9119

0.4114

0.0171

N,2-diphenylpyrido[2,3-d]pyrimidin-4-amine

31

0

2.5492

0.0000

0.0151

2,3,4,9-tetrahydro-1H-pyrido[3,4-b]indole

29

0

2.5492

0.0000

0.0141

(R)-2-benzyl-1-phenyl-2,3,4,9-tetrahydro-1H-pyrido[3,4-b]indole

27

0

2.5492

0.0000

0.0131

N-benzyl-2-(2-phenyl-2H-tetrazol-5-yl)aniline

41

5

2.2721

0.1789

0.0131

Pyrido[2,3-d]pyrimidine

32

2

2.3992

0.0968

0.0122

  1. aCount1 denotes the number of inhibitors containing the fragment
  2. bCount2 denotes the number of non-inhibitors containing the fragment
  3. cF1 denotes the fragment frequency in the inhibitor class
  4. dF2 denotes the fragment frequency in the non-inhibitor class