Prediction of Yields in Glucosylation of Rgocalciferol Through Response Surface Methodology and Artificial Neural Network Analysis.

-Glucosidase from sweet almond catalyzed synthesis of 20-O-(D-glucopyranosyl) ergocalciferol was analyzed using response surface methodology (RSM) and artificial neural network (ANN) analysis. In RSM a central composite rotatable design involving 30 experiments with four variables at five levels was employed to develop a predictive equation. The variables employed were immobilized -glucosidase 20 - 100% w/w of D-glucose, pH 4.0 - 8.0, buffer concentration 0.04 - 0.2 mM (0.4 - 2 ml) and incubation period 10 - 50 h. Surface plots clearly brought out the behaviour of immobilized -glucosidase in the glucosylation of ergocalciferol in exhibiting a crossover point at 60% (w/w D-glucose) immobilized -glucosidase, 0.1 mM buffer concentration and pH 6.0. A maximum yield predicted by RSM of 26% under 100% (w/w D-glucose) immobilized -glucosidase, pH 6.0, 0.12 mM (1.2 ml) buffer concentration and 30 h incubation period corresponded with an experimental yield of 26% under the same conditions. ANN analysis carried out with a set of training data also showed good correspondence to the testing data.