Skip to main content

Table 2 Classification results of the baseline models, MLP with NS, and MLP with PLANS using GINFP for the representation of chemical structures

From: Exploration of chemical space with partial labeled noisy student self-training and self-supervised graph embedding

  

Accuracy

Precision

Recall

F1

Cyp450

SVM

58.19 ± 0.81

0.68 ± 0.01

0.77 ± 0.01

0.72 ± 0.01

 

RF

54.38 ± 0.78

0.46 ± 0.01

0.79 ± 0.02

0.58 ± 0.01

 

AdaBoost

48.13 ± 3.86

0.23 ± 0.09

0.83 ± 0.14

0.35 ± 0.07

 

XGBoost

54.93 ± 0.78

0.59 ± 0.02

0.76 ± 0.01

0.66 ± 0.01

 

MLP

57.31 ± 1.47

0.74 ± 0.02

0.75 ± 0.02

0.74 ± 0.01

 

MLP + mixup

57.42 ± 0.46

0.65 ± 0.03

0.78 ± 0.02

0.71 ± 0.01

 

MLP + NS

59.83 ± 0.41

0.70 ± 0.02

0.79 ± 0.01

0.74 ± 0.01

 

MLP + mixup + NS

58.50 ± 0.48

0.65 ± 0.00

0.78 ± 0.01

0.71 ± 0.01

 

MLP + PLANS

60.61 ± 1.00

0.77 ± 0.01

0.79 ± 0.01

0.78 ± 0.01

 

MLP + PLANS + mixup

59.95 ± 1.41

0.72 ± 0.02

0.79 ± 0.01

0.75 ± 0.01

 

MLP + PLANS + balancing

61.58 ± 0.87

0.75 ± 0.02

0.80 ± 0.02

0.78 ± 0.01

 

MLP + PLANS + balancing + mixup

60.58 ± 1.38

0.73 ± 0.02

0.78 ± 0.02

0.76 ± 0.01

  

AP

F1

Tox21

MLP

0.04 ± 0.01

–

 

MLP + mixup

0.15 ± 0.07

0.01 ± 0.005

 

MLP + NS

0.08 ± 0.01

–

 

MLP + mixup + NS

0.14 ± 0.006

0.07 ± 0.03

 

MLP + PLANS

0.23 ± 0.02

0.24 ± 0.03

 

MLP + PLANS + mixup

0.16 ± 0.01

0.25 ± 0.01

 

MLP + PLANS + balancing

0.23 ± 0.02

0.23 ± 0.05

 

MLP + PLANS + balancing + mixup

0.16 ± 0.01

0.25 ± 0.02

  1.  The evaluation metric of the best performed model is highlighted in bold. The upper part shows the results for the CYP450 dataset and the lower part shows the results for the Tox21 dataset