A novel approach for the optimization of Shatavari (Asparagus racemosus Willd.) plant-based low alcohol nutra beverage production using Saccharomyces cerevisiae (NCIM 2428) in conjunction with artificial neural network and genetic algorithm (ANN-GA).

INTRODUCTION: This study focused on optimizing the fermentation conditions of Shatavari plant-based roots using an artificial neural network and response surface methodology. The aim was to identify the optimal independent variables and corresponding responses by comparing experimental and predicted responses. The experimentation was validated using a genetic algorithm, determining the best temperature, pH, and inoculum parameters. MATERIAL AND METHODS: In this study, we used the Shatavari (Asparagus racemosus Willd.) plant's root as their primary raw material and subjected it to treatment with α amylase and gluco-amylase enzyme (EC 232-885-6) which exhibited a remarkable activity level ranging from 8000 to 12,000 U/mg The resulting hydrolysate was fermented using Saccharomyces cerevisiae (NCIM 2428) culture. To determine the optimal combination of input variables a Central Composite Rotatable Design was implemented, facilitated by the Design Expert software (Version 11.0.3.0 by Stat-Ease Inc.),. RESULT AND CONCLUSION: The optimal conditions for the experiment were found to be a temperature of 32 °C, pH of 4.0, and inoculum concentration of 10% (v/v). The Artificial Neural Network (ANN) model was able to successfully predict the response variables with a marginal relative error rate of 8.722% and 24.312% for ethanol production and antioxidant activity, respectively. The fermented Shatavari-based low-alcohol Nutra beverage contained only fructose. The validation of Shatavari juice using the ANN model showed an enhanced ethanol yield of 3.21% and 421.47 μg/L antioxidant activity during fermentation. The experimental and predicted outcomes from the Artificial Neural Network-Genetic Algorithm (ANN-GA) model matched, proving its predictive precision. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s13197-025-06275-2.