Skip to main content

Table 3 The performance of models developed on EGFR10 dataset, class-specific molecules and EGFR10 excluding single class, evaluated using cross-validation techniques for testing on same-class of molecules

From: QSAR based model for discriminating EGFR inhibitors and non-inhibitors using Random forest

Trained on Tested on Sensitivity Specificity Accuracy MCC ROC
EGFR10 train EGFR10 train 68.74 87.67 84.95 0.49 0.89
EGFR10 train EGFR10 Validation 69.89 86.03 83.66 0.49 0.89
Pyrimidine Pyrimidine 69.25 92.13 86.92 0.62 0.92
Pyrimidine Quinazoline 68.62 54.88 58.88 0.21 0.67
Quinazoline Quinazoline 68.15 79.63 76.31 0.45 0.81
Quinazoline Pyrimidine 67.86 64.04 64.91 0.27 0.74
EFGR10-Pyrimidine EFGR10-Pyrimidine 68.7 94.08 91.34 0.59 0.92
EFGR10-Quinazoline EFGR10- Quinazoline 69.66 96.4 94.04 0.64 0.95
EFGR10- Pyrimidine Pyrimidine 68.06 76.74 74.77 0.4 0.77
EFGR10- Quinazoline Quinazoline 60.31 76.25 71.66 0.35 0.72