Valorization of Camel Milk Residue (CMR) into Hypoglycemic Peptides: An RSM-ANN Modeling Approach.

This study valorized camel milk residue (CMR) via optimized bacterial fermentation to produce bioactive peptides with hypoglycemic potential. Screening of eleven bacterial strains identified four optimal starters. Artificial neural network (ANN) simulation significantly outperformed response surface methodology (RSM) in modeling and prediction, as evidenced by its superior performance in key statistical metrics, including R, RMSE, and AAD(%), ultimately achieving a maximized yield of TCA-soluble nitrogen (TCA). Under optimized conditions, a TCA yield of 39.8% was achieved and experimentally validated. Ultrafiltration yielded a highly bioactive peptide fraction (<1 kDa), which exhibited significant inhibition of -amylase (80.7%) and -glucosidase (32.0%). The peptides exhibited high stability under various conditions, highlighting their industrial potential. This study explores the application of ANN-RSM optimization for the valorization of camel milk residue (CMR). Our findings provide a sustainable strategy for transforming CMR into a high-value anti-diabetic ingredient, which could contribute to extending the camel milk value chain.