摘要
为了探究氨基酸组成对β-琼胶酶催化底物时的最适温度的影响,分别统计分析了酶全长、N端、M段及C端序列中20种氨基酸出现的频次。以支持向量机回归构建了氨基酸组成与β-琼胶酶最适温度的模型,10-倍交叉验证获得最小平均绝对误差(MAE),线性核函数的M端为4.36℃,RBF核函数的M端为4.16℃,独立样本测试得到N端的MAE分别为3.12℃和4.43℃。上述结果表明,该酶N端和M端是影响其最适温度的重要因素。通过比较最适温度为60℃和30℃β-琼胶酶的高级结构,推测M端和N端中3个螺旋结构是影响该酶最适温度的重要因素。
To explore how the composition of amino acids influence the optimum temperature ofβ-agarases, we counted and calculated the frequences the amino acids for full-length, N terminal, M section and C-terminal sequence ofβ-agarases respectively. And then, we constructed support vector regression model to represent the relationships between amino acid composition and the optimum temperature. Based on the mean absolute error （MAE） of 10-fold cross-validation, the M section of linear and RBF kernel is the best, with the MAEs of 4.36℃ and 4.16℃, respectively. While the the minimum MAE of the independent test was obtained by the N terminal, with the MAEs of 3.12℃ and 4.43℃ respectively. The results showed that N terminal and M section ofβ- agarases might have great influence on its optimum temperature. By comparing the 3D structure of theβ-agarases with the optimum temperature of 30℃ and 60℃, we speculated that there were 3 helical structure differences in N terminal and M section, which should be the key factors that affect the thermostability of theβ- agarases.
出处
《计算机与应用化学》
CAS
2015年第12期1513-1518,共6页
Computers and Applied Chemistry
基金
国际海域资源调查与开发"十二五"重大项目(DY125-15-T-06)
国家高技术研究发展计划(2012AA092103)
福建省自然科学基金资助项目(2013J01048)
华侨大学研究生创新能力培养项目资金
关键词
Β-琼胶酶
氨基酸组成
最适温度
支持向量机回归
β-agarases
composition of amino acids
optimum temperature
support vector regression
