Optimization of Lyophilized Protective Agent Formulation of Lactobacillus casei LTL1361 Based on Artificial Neural Network Coupled Genetic Algorithm (BP-GA)

To improve the freeze-drying survival rate of Lactobacillus casei LTL1361 in the vacuum freeze-drying process, the single factor and Plackett-Burman designs were first conducted to verify the main factors affecting the freeze-drying survival rate of Lactobacillus casei LTL1361. According to the experimental results, the artificial network coupling genetic algorithm was constructed using the Box Behnken design. The artificial network coupled genetic algorithm (BP-GA) model was constructed to simulate and predict the lyophilized protective agent formulation of Lactobacillus casei LTL1361. The results showed that the three main factors affecting the lyophilisation survival rate of the strain were: Alginate, glutamic acid and mannitol, which were selected by single-factor and Plackett-Burman tests, and the three factors and the base skim milk were identified as the optimisation conditions for subsequent optimisation tests. The BP-GA model was used to find the optimum concentration of protectant for Lactobacillus casei LTL1361, which was 10.3% skim milk, 0.8% glutamic acid, 6.7% alginate and 4.0% mannitol, and the maximum lyophilisation survival rate of the strain was 89.56%. Using the BP-GA model, this study explored a probiotic lyophilisation protectant formulation with a high lyophilisation survival rate, and provided a reference for the preparation of lyophilised formulations of highly active strains and the development of commercial direct-injection ferments.