Computational modelling studies of some 1,3-thiazine derivatives as anti-influenza inhibitors targeting H1N1 neuraminidase via 2D-QSAR, 3D-QSAR, molecular docking, and ADMET predictions

Beni-Suef University Journal of Basic and Applied Sciences

Table 2 Internal validation of the 2D-QSAR models

Internal validation metrics	GFA-MLR model	GFA-ANN (4-5-1) model	Threshold	Comment	References
Lack of fit (LOF)	0.0546	–
Pearson correlation (r)	0.9590	0.9610	R > 0.6	Passed	[37]
Pearson correlation squared (\(R_{{\text{train }}}^{2}\))	0.9192	0.9227	R²_train > 0.6	Passed	[37]
Adjusted R² (R²_adj)	0.8991	–	R²_adj > 0.6	Passed	[37]
Spearman rank correlation (ρ)	0.9155	0.9220	ρ > 0.6	Passed
Root-mean-square error (RMSE)	0.0959	0.0940	Low	Passed
Cross-validated squared (Q²)	0.8767	0.9212	Q² > 0.6	Passed	[37]
Y-randomization (\(cR_{p}^{2} )\)	0.8300	–	\(cR_{p}^{2}\) > 0.6	Passed	[38]