结合蛋白质二级结构信息预测蛋白质空间结构中的二硫键*

在蛋白质空间结构预测中，二硫键的确定可以大大减少蛋白质构象的搜索空间。为提高二硫键预测的准确率，对形成二硫键的半胱氨酸及其周围的氨基酸残基在蛋白质二级结构形成上的偏性进行了分析，并提出将蛋白质二级结构信息加入到BP神经网络预测模型的输入编码信息中。研究对象为从SWISS-PROT数据库中选取的252条蛋白质序列，随机均分4组，对预测准确率进行4-交叉验证。各项准确率均比未加入蛋白质二级结构信息前，有明显提高。结果表明，结合蛋白质二级结构信息的编码方式是可行且有效的。

Disulfide bond prediction integrated protein secondary structure information

n protein-folding prediction, the location of disulfide bonds can strongly reduce the search in the conformational space. In order to improve prediction accuracy, analyzed the bias of free cysteines and cystines in the secondary structure preference, and proposed adding protein secondary structure information to the inputs of BP neural network. Selected 252 proteins from the SWISS-PROT database, and divided them into 4 even subsets. From the results of 4-fold cross validation, found that integrating protein secondary structure information can improve prediction accuracy. The results show that this encoding is feasible and effective.