Table 2 Standard validation parameters for a good QSAR model
From: Computational evaluation of some compounds as potential anti-breast cancer agents
Validation parameters | Meaning | Values |
|---|---|---|
\(R^{2}\) | Coefficient of determination | \(\ge 0.6\) |
\(P_{95\% }\) | Confidence interval at 95% confidence level | \(<\) 0.06 |
\(Q_{{{\text{cv}}}}^{2}\) | Cross-validation coefficient | \(>\) 0.5 |
\(R^{2} - Q_{{{\text{cv}}}}^{2}\) | Difference between \(R^{2}\) and \(Q_{cv}^{2}\) | \(\le\) 0.3 |
\(N_{{{\text{ext}}{.}\,{\text{test}}\,{\text{set}}{.}}}\) | Minimum number of external test sets | \(\ge 5\) |
\(R_{{{\text{test}}}}^{2}\) | Coefficient of determination for external test set | \(\ge\) 0.6 |
\({\text{cR}}_{{\text{p}}}^{2}\) | Coefficient of determination for Y-randomization | \(>\) 0.5 |