The best model was selected using model selection procedure (AICc).
Out of the models generated, the best model was selected based on Modeller's scoring function and external validation via Whatcheck (Hooft et al., 1996) and PROSA (Wiederstein and Sippl, 2007).
The best model was selected between those that perform feature selection, i.e. GLM, LASSO, SVM-RFE, RF-RFE, ENET.
The best model was selected and analyzed based on Modeller's probability density function (Discrete Optimized Protein Energy score) and the Ramachandran plot (Laskowski et al., 1993), and subsequently refined using the Loop refine module of Modeller.
The best model was selected according to the Akaike's and Bayesian Information Criteria.
The best model was selected based on overall stereochemical quality (Procheck, PROSA, GenThreader), energy minimized, refined and used for active site characterization in BioMed CAChe workspace.
A 90% confidence set of models was compiled if no best model was selected via AICc.
The influences of grid distances, alignment methods and combination of charges were explored out of which the best model was selected.
Then, the best model was selected and the 2000 and 2006 maps (Figs. 10, 11) simulated and Percent Area Match location and quantity have been used for evaluating the goodness of fit the simulated maps (Tables 7, 8).
The best model was selected based on the comparison of the area under the curve generated by each model, where the size of the area reflects the amount of information included in the calculations.
The best model was selected based on the highest coefficients of determination for calibration and cross-validation (( {text{R}}_{text{cal}}^{2},;{text{R}}_{text{cval}}^{2} ),) and the smallest standard error of calibration (SEC), standard error of cross validation (SECV) or standard error of prediction (SEP) and the lesser number of latent variables used.
