多因素分析对肺吸虫囊蚴脱囊优化条件的探讨

Multiple factors analysis in discussing the optimum condition of the Paragonimus westermani cysts decysted

目的对正交设计试验结果进行多因素分析,确定肺吸虫囊蚴脱囊影响因素,选出脱囊条件的最优水平搭配;建立多元线性回归模型,分析比较各影响因素作用性质和大小及影响趋势。方法1)根据专业知识,选定影响肺吸虫囊蚴脱囊的4个主要因素,每个因素取4个水平。选用L16(45)正交表安排试验;2)对正交设计试验结果进行方差分析;3)对正交设计试验资料进行回归诊断,建立多元线性回归模型,分析比较各影响因素作用性质和大小及影响趋势。结果试验结果分析得知,影响肺吸虫囊蚴脱囊的显著因素是pH值为2的胃酶浓度、胃酶作用时间和胆盐浓度,胰酶影响作用不显著。对正交设计试验资料回归诊断,发现第一组试验资料所对应的点为异常点,删去异常点,建立多元线性回归一般方程(y^=22.89536-99.19433x1-0.23107x2-2.49086x3-69.48004x4)。随着胃酶、胰酶及胆盐浓度升高,胃酶作用时间延长,囊蚴脱囊数量减少。肺吸虫囊蚴脱囊条件最优搭配为:胃酶浓度为0.025%(pH=2),作用时间为10min,胆盐、胰酶液分别为0.025%和0.25%。胆盐浓度为0.025%并且pH值为2的条件下胃酶浓度对肺吸虫囊蚴脱囊影响作用最大,其次为胆盐浓度、胃酶作用时间,而胰酶浓度对肺吸虫囊蚴脱囊影响作用最小。结论正交试验的方差分析与多元线性回归方程分析相结合,提高了试验数据的利用效率,扩大了试验结果的分析范围,在相同的投入下得出了更多的信息,提高了试验效率。本研究的思路和方法为同类研究提供了有益的启示和借鉴。

Objective The influential factors of the Paragonimus westermani cysts decysted were confirmed and the optimum arrangement in pairs or groups of influential factors was found out through multiple factors analysis of orthogonal experimental results in this study. Multiple linear regression model was established, the character and the action of each independent variable was analyzed and compared. Methods 1）According to professional knowledge, four main influential factors were chosen, each of which had four levers. The test was performed according to the orthogonal table L16 （4^5） ;2）Analysis of variance （ANOVA） was used to deal with the orthogonal test results;3）A regression diagnosis was made to the orthogonal data,The multiple liner regression model was established. The action of each influential factor of the P. westermani cysts decyst was analyzed and compared. Results Based on multiple factors analysis, pepsin density （pH= 2）, pepsin action time and bile salt density are statistically significant to P. westermani cysts decyst, but pancreatin density is not statistically significant. The optimum arrangement in pairs or groups of influential factors is that： pepsin density （pH= 2） is 0. 025 % ; pepsin action time is 10 minutes; and bile salt density is 0. 025 %. Through a regression diagnosis, the first group data of the orthogonal test is an outlier. Deleted the outlier, the multiple linear regression equation is： y^-= 22. 89536-99. 19433x1 -0.23107x2- 2. 49086x3 -69. 48004x4. The amounts of P. westermani cysts decysted will reduce when the pepsin density, bile salt density and pancreatin density increase, or the pepsin action time prolong. And the pepsin density is the strongest influential factor, the bile salt density is the stronger one, pepsin action time is the third one and pancreatin density is the last one to P. westermani cysts decyst. Conclusion Combined the ANOVA of orthogonal test with the analysis of multiple linear regression equation, the test data could be made full use of, the analysis scope of experimental results is widen, and more information is acquired under the same condition. The method of this study can offer helpful reference and inspiration to the similar research.