In silico screening for identification of novel β-1,3-glucan synthase inhibitors using pharmacophore and 3D-QSAR methodologies

SpringerPlus

Table 5 Summary of PHASE 3D-QSAR statistical results of AAARR.594 hypothesis, best model is shown in italic font

No. of factors^a	SD^b	R-squared^c	F^d	P^e	Stability^f	RMSE^g	Q-squared^h	Pearson-Rⁱ	\(R_{pred}^{2}\) ^j
1	0.3265	0.7089	70.60	2.909e−009	0.7744	0.2226	0.7318	0.8101	0.656
2	0.2281	0.8628	88.10	8.336e−013	0.7296	0.2170	0.7450	0.8900	0.698
3	0.1613	0.9339	127.20	4.879e−016	0.5590	0.2015	0.7801	0.9011	0.689
4	0.1368	0. 9542	135.30	5.277e−017	0.4923	0.1953	0.8268	0.9103	0.711

^aNumber of factor used in the analyses
^b(SD) the standard deviation of regression
^c(R ²) coefficient of determination
^d(F) the ratio of the model variance to the observed activity variance
^e(P) significance level of F when treated as a ratio of Chi squared distributions
^f(Stability) stability of the model predictions to changes in the training set composition
^g(RMSE) the RMS error in the test set predictions
^h(Q ²) directly analogous to R ² but based on the test set predictions
ⁱ(Pearson-R) value for the correlation between the predicted and observed activity for the test set
^j(\(R_{pred}^{2}\)) Predictive R ², Standard deviation of error prediction