. Comparisons between model predictions and measured values for available lysine losses in a model food system.

A mathematical model, developed to predict available‐lysine losses in a model food system which had undergone an isothermal‐nonstirred process, was tested under different process conditions. The model predictions were compared to measured available‐lysine losses in both jacketed‐mixer processing and extrusion processing. Statistical analysis using a paired t‐test analysis indicated a significant correlation between predicted and actual results for the jacket‐mixer process. In the extrusion experiment a statistically significant correlation (0.7) occurred when the predicted values were plotted against observed. However, variation due to product backmixing and inaccurate temperature measurement prevented a rigorous test of the model. The reaction order of available lysine loss in casein, single‐cell protein and a soy protein isolate were determined at an elevated temperature. The casein and single‐cell protein followed first‐order reaction kinetics. The soy isolate initially follows first‐order loss but after an approximate loss of 40–50% available lysine, the loss abruptly stops and a no‐loss phase occurs. Copyright © 1978, Wiley Blackwell. All rights reserved