摘要
培养基优化属多变量目标函数优化,先通过Plackett-Burman法筛选出黄嘌呤、(NH_4)_2SO_4和CaCl_2为重要因素,然后采用Box-Behnken设计重要因素水平,同时应用SASV8.1软件做多元非线性回归,建立黄嘌呤氧化酶产酶培养基的优化模型。用模型预测培养基的最优组成。将预测值与验证值相比较,误差为2.6%,证明采用响应面法寻求最佳培养基组成是可行的;并且优化后黄嘌呤氧化酶产酶水平提高了5倍。
The optimium medium components for xanthine oxidase is to find the optimal solution for a nonlinear objective-function. SASV8.1 was used to optimize the medium components for xanthine oxidase. The effects of eight additives of the media were evaluated using a Plackett-Burman design. Results showed that xanthine, （NH4 ）2SO4 and CaCl2 of the media were the main effecting factors while other additives had no significant effect on the yield of xanthine oxidase. A regression model was obtained by fitting the experimental date to full second-order polynomial model using a Box-Behnken design. Optimal concentrations of the medium components was concluded finally. The prediction error is 2. 6% through experimental verification. The activity of xanthine oxidase was 4. 623 U/mL, which was 5 times more than the control.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2008年第5期611-614,共4页
Computers and Applied Chemistry
基金
工业生物技术教育部重点实验室基金(KLIB-KF200502)
