Skip to main content

Table 1 QSAR models validation parameters and scores

From: Computer-aided identification of a series of novel ligands showing high potency as hepatitis C virus NS3/4A protease inhibitors

Statistical parameter Values for the developed model Threshold value Comments Source
Internal validation (validation with the training set data)
 LOF 0.1574    
 R2 0.7704 > 0.6 Passed a
 R2adj 0.7402 > 0.6 Passed b
 F 25.4979    
 Q2LOO 0.6914 > 0.5 Passed b
 SEE 0.4175    
 RMSE 0.3880    
 \(R_{{m\left( {{\text{LOO}}} \right)}}^{2}\) 0.6507 > 0.5 Passed c
 \(R_{{m\left( {{\text{LOO}}} \right)}}^{2^{\prime}}\) 0.5106 > 0.5 Passed c
 \(\Delta R_{{m\left( {{\text{LOO}}} \right)}}^{2}\) 0.1401    
 cR2p 0.7025 > 0.5 Passed d
External validation (validation with test set data)
 R2(pred) 0.7047 > 0.6 Passed a
 r2 0.6359 > 0.5 Passed e
 r20 0.6235 > 0.5 Passed e
 r20 0.5651 > 0.5 Passed e
 |r20 − r20| 0.0584 < 0.3 Passed e
 \(r^{2} - r_{o}^{2} /r^{2}\) 0.0195 < 0.1 Passed f
 k 1.0004 0.85 < k < 1.15 Passed f
 \(r_{o}^{2} - r_{o}^{^{\prime}2} /r_{o}^{2}\) 0.0937 < 0.1 Passed f
 k 0.9979 0.85 < k < 1.15 Passed f
  1. a = (Veerasamy et al. 2011), b = (Tropsha 2010), c = (Damme and Bultinck 2007), d = (Roy et al. 2012), e = (Golbraikh and Tropsha 2002), f = (Tropsha et al. 2003)
\