Skip to main content

Table 4 Ligand-based and structure-based CoMSIA models along with percentage contribution of their descriptors

From: A combined 3D-QSAR and docking studies for the In-silicoprediction of HIV-protease inhibitors

COMBINATIONS q2 r2 r pred 2 F C SEE SEP 1% 2% 3% 4% 5%
S+ES 0.714 0.927   150.91 6 0.524 1.038 37.1 62.9 _ _ _
S+H 0.728 0.935   171.61 6 0.493 1.014 42.9 57.1 _ _ _
S+D 0.681 0.89   95.284 6 0.645 1.096 62.5 37.5 _ _ _
S+A 0.756 0.941   190.33 6 0.47 0.959 55 45 _ _ _
S+D+A 0.727 0.936   171.69 6 0.493 1.015 42 34.7 23.3 _ _
ES+H 0.728 0.935   171.28 6 0.494 1.014 53.2 46.8 _ _ _
ES+D 0.69 0.912   122.84 6 0.576 1.082 70.6 29.4 _ _ _
ES+A 0.755 0.945   203.06 6 0.456 0.96 62.1 37.9 _ _ _
ES+D+A 0.713 0.938   178.42 6 0.484 1.041 50.8 28.8 20.4 _ _
H+D 0.692 0.917   130.07 6 0.561 1.078 68.2 31.8 _ _ _
H+A 0.745 0.942   193.39 6 0.466 0.98 59.7 40.3 _ _ _
H+S+ES 0.731 0.946   206.42 6 0.452 1.007 35.1 24.8 40.2 _ _
H+D+A 0.717 0.934   167.58 6 0.499 1.033 47 30.9 22.1 _ _
D+A 0.493 0.825   55.777 6 0.812 1.383 41 59 _ _ _
D+S+ES 0.707 0.932   161.84 6 0.507 1.051 23.3 27.7 49 _ _
A+S+ES 0.755 0.952   232.8 6 0.427 0.962 29.1 26.9 44 _ _
S+ES+D+A 0.732 0.948   214.79 6 0.444 1.005 21.9 37.7 22.6 17.8 _
S+H+D 0.718 0.936   173.05 6 0.491 1.032 32.7 43.8 23.5 _ _
S+H+A 0.766 0.952   234.85 6 0.425 0.939 30.2 40 29.9 _ _
ES+H+D 0.727 0.94   185.88 6 0.475 1.014 44 35 21 _ _
ES+H+A 0.761 0.953   238.73 6 0.422 0.949 40.5 33.2 26.3 _ _
H+D+S+ES 0.733 0.948   216.44 6 0.442 1.003 28 18.6 19.8 33.6 _
H+A+S+ES 0.762 0.958   272.47 6 0.396 0.948 27.2 21.5 19.5 31.8 _
S+H+D+A 0.747 0.949   220.3 6 0.438 0.977 25.1 33.6 23.5 17.8 _
ES+H+D+A 0.747 0.954   246.39 6 0.416 0.977 34.9 27.4 21.4 16.2 _
H+S+ES+D+A 0.751 0.958 0.93 270.35 6 0.398 0.97 22.8 16.5 28.2 18.1 14.4
Structure Based Model
H+S+ES+D+A 0.664 0.955 0.927 252.32 6 0.411 1.125 26.2 15.4 25.4 16.7 16.3
  1. Where: q2: cross validated correlation coefficient, r2: non-cross validated correlation coefficient, r pred 2 : prediction of external test set for validation, F Fischer test values, C optimal number of Components, SEE Standard Error of Estimation, SEP Standard Error of Prediction, %1-5: percentage contribution of descriptors in the field, respectively, S Steric field, ES Electrostatic field, H Hydrophobic descriptor, D hydrogen bond Donor field, and A hydrogen bond Acceptor field.