Skip to main content

Table 2 Correlation values for MLR and SVM based QSAR models developed using descriptors selected at pair-wise correlation cut-off value 0.9

From: Ki DoQ: using docking based energy scores to develop ligand based model for predicting antibacterials

Number of input variables

R

q 2

MAE

Using MLR (23 inhibitors)

4

0.81

0.65

2.61

Using SVM (23 inhibitors)

8

0.87

0.75

2.24

7

0.91

0.81

2.01

6

0.90

0.81

2.16

5

0.90

0.79

2.19

Using SVM (20 inhibitors)

5

0.95

0.89

1.28