贝塔一葡萄糖苷酶催化反应的最适pH值和最适温度的同时预测

利用氨基酸特征同时预测贝塔一葡萄糖苷酶反应的最适pH值和最适温度.首先,用不同的氨基酸特征对贝塔一葡萄糖苷酶进行量化作为输入,以最适pH值和最适温度作为输出；然后用20-2前馈反向传播的神经网络进行训练；最后用子集验证、刀切验证和交叉验证三种方法进行验证.结果表明,在24个氨基酸特征中只有4个特征可以用于模型预测,其中...

Simultaneously Predicting Optimum pH Value and Optimum Temperature in Catalytic Reaction of Beta-glucosidase

The features of amino acids were used to simultaneously predict optimum pH value and optimum temperature of beta-glucosidases.Firstly,the beta-glucosidases were quantified by different features of amino acids as inputs,and their optimum pH value and optimum temperature were served as outputs;secondly,the training was conducted by means of 20-2 feedforward backpropagation neural network;finally,the validation was performed by three approaches,subset validation,jackknife validation,and cross-validation.The re...