论文部分内容阅读
基于氨基酸组成和有偏自相关函数的特征参量 ,利用BP神经网络 ,提出了一种预测蛋白质二级结构中α螺旋和 β折叠含量的计算方法 .采用相互独立的非同源蛋白质数据库对该方法的准确性进行检验 ,对蛋白质二级结构α螺旋和 β折叠含量的预测的结果为 :自检验的平均绝对误差分别为 0 .0 70和 0 .0 6 8,相应的标准偏差分别为 0 .0 49和 0 .0 47;他检验的平均绝对误差分别为 0 .0 75和 0 .0 70 ,相应的标准偏差分别为 0 .0 5 0和 0 .0 49.与常用方法相比 ,利用此方法预测蛋白质二级结构含量可有效提高预测精度 .
Based on the characteristic parameters of amino acid composition and biased autocorrelation function, a BP neural network was proposed to calculate the content of α-helix and β-sheet in the secondary structure of protein.An independent protein database was used to analyze this method The results of the prediction of α-helix and β-sheet content of secondary structure of protein showed that the average absolute errors of self-test were 0.70 and 0.608, respectively, and the corresponding standard deviations were 0 respectively. 0 49 and 0 0 47; he test the average absolute error of 0 .0 75 and 0 .0 70, respectively, the corresponding standard deviation of 0 .0 5 0 and 0 .0 49. Compared with the common method, the use of This method can predict the secondary structure of protein effectively and improve the prediction accuracy.