Optimization the preparation of whey protein from pasteurized liquid milk using Response Surface Methodology and Artificial Neural Network Model

To produce protein hydrolysates with antioxidant and amino acid activities, whey protein concentrate (WPC) was alcalase-acylated under the optimal conditions in this study using Response Surface Modeling (RSM) and Artificial Neural Network (ANN). Whey protein is a component of milk regarded as a beneficial nutrient. ANN model has been trained by three input neurons which represent the hydrolysis time, pH, and temperature, and four output neurons representing the yields (%), protein (%), α, α-diphenyl-β-picrylhydrazyl (DPPH) inhibition activity (%) and amino acids (%). The optimized hidden layer neurons were obtained based on a minimum mean squared error. A polynomial function was used to analyze the association between the variables and their individual and joint effects. Fourier-transform infrared spectroscopy (FTIR) analysis was done with Attenuated total reflection (ATR) mode. An optimization technique was done using Central Composite Design (CCD). The ideal Alcalase-hydrolysis settings were optimized for producing whey protein using RSM of DPPH (Association of Official Analytical Chemists (AOAC) method) and amino acid activity (High Performance Liquid Chromatography (HPLC) method). The Expected R 2 and the Adjusted R 2 are within a tolerable range, i.e., the variability is <0.2. The signal-to-noise ratio is measured with acceptable precision. The model's ability to generate the predicted answers with a high degree of confidence was suggested by the overall desirability (D-value) of 1.00 with solution one. According to the findings, the ideal conditions were reached at potential of hydrogen (pH)= 6.14, hydrolysis time= 4.00 h and temperature= 44.28 °C to produce the highest yields (8.55 %), highest levels of protein (9.46 %), highest levels of α, α-diphenyl-β-picrylhydrazyl (DPPH) inhibition activity (45.96 %) and highest levels of amino acids (7.00 %). The regression model's suitability was confirmed by the well-fitted analysis data. The present study showed that WPC could produce bio-functional hydrolysates with antioxidant and amino acid activity. This finding supports the use of WPC hydrolysate as a new natural ingredient for the production of functional food products.