目的 在大豆蛋白中导入酪蛋白，弥补大豆蛋白中Met含量过少的不足，并提高其功能。方法 以大豆蛋白和酪蛋白为原料，采用α-糜蛋白酶催化类蛋白反应，向大豆多肽（SP）中导入酪蛋白非磷酸肽（CNPP），形成大豆蛋白-酪蛋白重组体（CNPSPC），并对其产率进行测定。然后基于人工神经网络模型对CNPSPC反应条件进行优化。结果 当底物CNPP与SP的质量比为4∶1，反应温度为46.2 ℃，反应pH为4.97，反应时间为20 h时，CNPSPC的产率达到最大，可达（87.0±0.10）%。结论 该预测模型可以很好地对大豆蛋白-酪蛋白类蛋白反应的工艺条件进行优化，优化的CNPSPC可作为很好的生物膜原料。

The work aims to introduce casein into soy protein to make up for the deficiency of too little Met in soy protein and to improve the functional properties. Soybean protein and casein were used as raw material to introduce soybean peptide (SP) in the casein non phosphopeptides (CNPP) by chymotrypsin catalyzed plastein reaction so as to form soybean protein-casein complexes (CNPSPC), and determine the yield. Then the reaction conditions of CNPSPC were optimized based on the artificial neural network model. When the ratio of substrate addition was 4∶1 (CNPP and SP), the reaction temperature was 46.2 ℃, the reaction pH was 4.97, and the reaction time was 20 h, the yield of CNPSPC reached a maximum of (87.0±0.10)%. This prediction model can optimizer the process of soy protein-casein protein reaction better. The optimized CNPSPC can be used as a good biofilm material.