MLP Kernel-Based to Predict the Optimal Conditions of Transglutaminase on Protein Polymerization

The research used the MLP kernel-based numerical simulation and estimation method. To analyze the effect of transglutaminase (TGase) on goat milk protein. Find the optimum temperature, the optimum amount of TGase to add, and the optimum reaction time. The experimental method was to remove the fat by high-speed centrifugation at 10,000 xg, 20 min, and 4 °C in goat milk and add different contents of TGase (0, 0.25, 0.1, 0.2, and 0.3 g). Then, a sample with a content of TGase was reacted at different temperatures (30, 40, and 50 °C) and different reaction times (1, 2, and 3hr), and then SDS gel electrophoresis was performed. The cross-linking conditions at different concentrations were observed. After the reaction, SDS gel electrophoresis was carried out, and at the end, it was fixed with a fixative solution and stained with a staining solution, and finally, the results were obtained by destaining. To get the best translation repair, an artificial neural network based on the MLP kernel was used as the estimation engine. Using the experimental results as simulation data, parameters such as optimal experimental temperature, TGase dosage, and experimental duration were calculated. The experimental results showed that the optimal conditions were the strongest at 47 °C, 0.1 g/mL, and 1 hr, with TG, and had the greatest effect on casein in goat milk. TGase acts most deeply on $\kappa-casein$, making casein polymerization more pronounced. This was consistent with the estimation conclusion of the numerical simulation.