The Optimization Of The Practice Of QbD Used In The Extraction Process Of Dysosmaversipellis And Glycyrrhiza By SVR

The stability of design space determines the effect of the practice of QbD(Quality by Design)concept,and the fitting effect of predicting models has much influence on the stability of design space.Quadratic Polynomial Model(QPM)is the most used model in the process of QbD implement,whose limitation of prediction accuracy and application cope is the main reason accounting for the lack of stability of deisgn space.In this work,we aim at improving the stability of design space,and the specific research contents and results are as follows:1.The optimization of the extraction of podophyllotoxin from Dysosmaversipellis and Glycyrrhiza,and BPANN(back propagation artificial neural network)and SVR(support vector regression)were established and optimized respectively to fit the relationship bwtween process parameter and exctaction ratio of podophyllotoxin.As a result,the SVR showed superiority in predicting accuracy,stability,generalization ability and amount of data needed when modeling after a comparison.The amount of extractor added,the content of extractor and proportion between Dysosmaversipellis and Glycyrrhiza were identified as the critical process parameters,increasing the reliability.In addition,the design space was searched by models of QPM and SVR.After the verification and comparison of the accuracy and stability of design space,the stabilities of design space are 0.9509 and 0.9884 for QPM aand SVR,respectively.The better fitting effect of SVR led to a more stable design space.2.In order to improve the anti-cancer effect of Glycyrrhiza and Dysosmaversipellis,the compatibility ratio of Glycyrrhiza and Dysosmaversipellis was optimized based on the quality by design concept.The inhibition rate of crude extracts from Glycyrrhiza and Dysosmaversipellis on Hela cells was the pharmacodynamic evaluation index.After an optimization and validation of the design space,the final drug proportion design space was [4.55,8.15].We introduced SVR model with better fitting effects into the implementation of QbD increasing the accuracy and reliability of technological parameters,which is proved to be an effective method to improve the stability of design space.We also optimized the compatibility ratio of Glycyrrhiza and Dysosmaversipellis and improved their antitumor activity by this way.