人工神经网络优化酶法制备马氏珠母贝低聚肽的工艺研究

以马氏珠母贝全脏器为原料,采用胃蛋白酶、胰蛋白酶以及胰凝乳蛋白酶模拟消化道酶解的方式制备低聚肽,先通过胃蛋白酶酶解获得初产物,在此基础上利用人工神经网络对胰蛋白酶及胰凝乳蛋白酶的双酶酶解过程进行模拟预测,对其酶解工艺进行优化.试验结果表明,通过人工神经网络预测获得小肠消化蛋白酶双酶酶解的优化工艺条件为:复合酶添加量4 500 U/g(m 胰蛋白酶∶m胰凝乳蛋自酶=6∶1)、料水比1∶2(W/V)、酶解时间3.5 h,在此条件下制备的低聚肽(200～2 000 u)得率达55％以上.通过优化的工艺条件制备的低聚肽可获得较高的得率,能够为马氏珠母贝全脏器低聚肽的开发利用奠定基础.

Optimization of hydrolysis conditions for productionof oligosaccharide from Pinctada martensii usingartificial neural networks

In this research, Pinctada martensiiwas was used as the raw material, as its oligosaccharide was obtained bysimulating the conditions of gastrointestinal tract digestive system. After hydrolyzed with pepsin, the hydrolysis process wasoptimized under the combined action of trypsin and chymotrypsin. According to the analysis of artificial neural networks,the result demonstrated that enzyme-to-substrate level of 4 500 U/g (trypsin/chymotrypsin ratio of 61), substrate-to-liquidratio of 12 and reaction time of 3.5 h were the optimal conditions to generate the oligosaccharide (200-2 000 u MW).Under these conditions, a relatively high yield of P. martensii oligosaccharide could be obtained, which would lay afoundation for its further exploration and utilization.