Study On The “Four Natures” Of Pfaffia Paniculata

Objective:To explore the scientific nature of the drug properties of Pfaffia paniculata,C5.0 algorithm and C&R classification regression algorithm were used to construct a decision tree for the cold and heat properties of traditional Chinese medicine;use non-targeted metabolomics to detect endogenous metabolites in the urine of rats in the drug group combined with the Orthogonal partial least squares discriminant analysis method to establish cold and heat The medicinal properties prediction model which was used to predict the cold and hot medicinal properties of Pfaffia paniculata.Urine metabolomics analysis were used to further analyze the metabolic mechanism of classification of cold and heat nature to explore the scientific nature of Pfaffia paniculata’s medicinal properties.Mtthods:1 Construction of the “four natures” Decision Tree Based on 6 Typical Cold and Heat Medicines——Prediction of the “four natures”of Pfaffia paniculataThe rats were randomly divided into blank control group,hot medicine group(aconite group,dried ginger group,cinnamon group,pepper group,evodia group,galangal group),cold medicine group(scutellaria baicalensis,coptis group,phellodendri group,gentian group),Gardenia group,Sophora flavescens group),each group was given the corresponding medicine decoction by gavage,the gavage volume was 10ml/kg,and the administration was administered for a total of 29 days.The number of spontaneous activities of the rats was detected.The rats were sacrificed and samples to be tested were collected.Energy-related indicators include liver and skeletal muscle Na~+K~+-ATPase activity,liver and skeletal muscle SDH enzyme activity,liver and skeletal muscle glycogen content,serum TG,CHOL,HDL,LDL,NEFA content and body fat coefficient;Endocrine system indicators include testis coefficient,adrenal coefficient,serum T4,T3,TSH,T and urine 17-hydroxycorticosteroid content;blood routine and liver and kidney function related indicators ALT,AST,AKP,TP,AKP,ALB,A/G,CK,UREA,CR,UA.Summarize the test results of all indicators and import them into the SPSS modler 14.0 software.Use the blank group,hot medicine group,and cold medicine group as the training set to construct the C5.0 decision tree and C&R decision tree,and then use the blank group,hot medicine group,and cold medicine group as the Testing set.The group’s data was used to test the model and to determine whether the model was reliable before predicting the medicinal properties of Pfaffia paniculata.2 Based on orthogonal partial least squares discriminant analysis method to establish the “four natures” prediction model of medicinal properties—prediction of Pfaffia paniculata “four natures”The rats were randomly divided into heat medicine group(aconite group,dried ginger group,cinnamon group,prickly ash group,evodia group,galangal group),cold medicine group(scutellaria baicalensis,coptis group,phellodendron,gentian group,gardenia group),Sophora flavescens group),blank control group,Pfaffia paniculata group,each administration group was given the corresponding water extract by intragastric administration,and the 12 h urine of each group was collected on the first 1,8,15,22,and 29 days after administration.QTOF/LC-MS LCMS system collects rat urine metabolomics signals.The data was processed by SIMCA 14.1multivariate statistical analysis software,and the mean value of metabolomics data of each group of rats at different time points was analyzed by PCA,and the time point with the largest difference in metabolism was screened out,which was further based on orthogonal partial least squares analysis(OPLS-DA)Establish a model for discriminating the cold and hot properties of traditional Chinese medicine,and import the data of Pfaffia paniculata into the established model for predictive analysis of the properties.3 Based on the urine metabolomics method to explore the typical cold and heat medicines and the "Four Qi" classification mechanism of Pfaffia paniculataStudy on the classification mechanism of cold and heat properties of traditional Chinese medicine and the attributes of Pfaffia paniculata based on urine metabonomics.After 7 days of adaptive feeding,the rats were randomly divided into a hot medicine group(aconite group,dried ginger group,cinnamon group,pepper group,evodia group,galangal group),cold medicine group(scutellaria group,coptis group,phellodendri group,Gentian group,Gardenia group,Sophora flavescens group),blank control group,Pfaffia paniculata group,each administration group was given the corresponding water extract by intragastric administration,and each group was collected on the 1,8,15,22,29 days of administration 12 h urine,UPLC-QTOF/LC-MS system was used to collect rat urine metabolomics signals.The data was processed by SIMCA 14.1 multivariate statistical analysis software,and the metabolites of each group of rats at different time points were processed PCA analysis of the average value of the scientific data,the time point with the largest difference in metabolism was screened,and the OPLS-DA multivariate statistical analysis method was used to calculate the difference metabolite of VIP>1 between the hot medicine group and the blank group,and the cold medicine group and the blank group.Potential biomarkers,using Progenesis QI software combined with HMDB human metabolite database to identify endogenous biomarkers,Metabo Analyst 5.0 software combined with KEGG database for biomarker enrichment analysis and metabolic pathway analysis,to explore the classification of “four natures”of traditional Chinese medicine Mechanism and molecular mechanism of the cold and heat properties of Pfaffia paniculata.Results:1 Construction of the “four natures” Decision Tree Based on 6 Typical Cold and Heat Medicines——Prediction of the “four natures” of Pfaffia paniculataThere are 8 branches in the decision tree constructed using the C5.0 algorithm,and the accuracy of the model test is 98.18%.The analysis of the importance of the algorithm’s variables shows: in the C5.0 decision tree,the importance of the contribution to the discrimination of cold and hot medicine properties The indicators of RDW,glycogen content in liver tissue,serum FFA content,and serum SDH enzyme activity;The decision tree constructed by the C&R classification regression algorithm has 4 branches,and the model test accuracy rate is 88.18%.The importance of algorithm variables The analysis results show that in the C&R classification regression algorithm,the importance indexes of the contribution to the discrimination of cold and heat drugs are as follows: FFA content in serum,glycogen content in liver tissue,RDW,MC in blood routine index,T4 and TSH content in serum,SDH enzyme activity in skeletal muscle tissue,testicular coefficient,Lymph and WBC indicators in blood routine.Import Pfaffia paniculata data into C5.0decision tree and C&R decision tree: C5.0 decision tree predictive classification results show that the probability that 8 samples of Pfaffia paniculata(n=10)are judged as hot is 95.70%,and two samples are judged as cold The C&R decision tree prediction classification results show that the probability of eight samples of Pfaffia paniculata(n=10)as hot is 94.10%,and the probability of the other two samples as cold is 77.60% and 46.20%,respectively.2 Based on orthogonal partial least squares discriminant analysis method to establish the “four natures” prediction model of medicinal properties—prediction of Pfaffia paniculata “four natures”Metabonomics data collected in the positive ion mode Chinese heat medicine group(Aconite group,dried ginger group,cinnamon group,pepper group,evodia group,galangal group),cold medicine group(Scutellaria group,Coptis group,Phellodendron chinense group,Gentian group,gardenia group,sophora flavescens group)PCA pattern recognition,suggesting that there is a classification trend between the two groups,the difference within the group is small.The metabolomics data collected in the positive ion mode is modeled in hot medicine include Aconite,dried ginger,cinnamon,evodia,Zanthoxylum bungeanum and cold medicine of Scutellaria,Coptis,Phellodendron,Gardenia and Sophora flavescens as the traning set,when galangal and gentian are used as the test set,the model has good training ability [R~2X(Cum)=0.521,R~2Y(Cum)=0.943)and test capability [Q2(Cum)=0.82],the accuracy rate of the model test is 93.75%;The metabolomics data collected in the negative ion mode is modeled with aconite,dried ginger,cinnamon,evodia,Zanthoxylum bungeanum,and scutellaria,coptis,gentian,gardenia,and sophora flavescens in the cold medicine as the training set,using galangal and cork as the test set,Test set the model has good training ability [R~2X(Cum)=0.455,R~2Y(Cum)=0.985,] and test ability [Q2(Cum)=0.829],the model test accuracy rate is 100%.The cold and heat prediction model established by the Pfaffia paniculata(n=8)positive and negative ion model has consistent prediction results.Seven samples are predicted to be hot and one sample is predicted to be cold.3 Based on the urine metabolomics method to explore the typical cold and heat medicines and the“four natures” classification mechanism of Pfaffia paniculataIn this study,the gardenia group was administered on the 8th day,the phellodendri group was administered on the 15 th day,the aconite group and the Pfaffia paniculata group were administered on the 29 th day,the dried ginger group,galangal group,pepper group,cinnamon group,evodia group,dragon The urine metabolomics data of the gall bladder group,sophora flavescens group,scutellaria group,and coptis group on the 22 nd day of administration were analyzed.The results showed that the typical cold and heat drugs in the negative ion mode mainly affected 26 biomarkers such as galactonic acid,glyceraldehyde,2,3-dihydroxy-2-methyl-propionic acid,etc.The adjustment of the content of markers basically has the opposite trend,and the adjustment trends of Pfaffia paniculata and thermal medicine on biomarkers are basically the same.Typical cold and heat medicines are ascorbic acid and uronic acid metabolism,pentose and glucuronic acid mutual conversion,pentose phosphate pathway,fatty acid elongation,fatty acid degradation,and tyrosine metabolism regulation are in opposite directions.Pfaffia paniculata and heat medicine regulate metabolic pathways The same direction.Conclusion:In this study,a reliable C5.0 decision tree and C&R decision tree for the prediction of the "four qi" of traditional Chinese medicine were constructed based on the effects of typical cold and heat drugs on the physiological and biochemical indicators of rats;based on the effect of typical cold and heat drugs on the endogenous metabolism of rat urine A predictive model of cold and heat properties under the positive and negative ion mode was established.The above decision tree and model show that Pfaffia paniculata is a hot medicine.Metabonomics studies have found that typical cold-heat drugs regulate the metabolism of ascorbic acid and aldonic acid,glutaric acid,and glutaric acid by regulating 26 biomarkers such as galactonic acid,glyceraldehyde,2,3-dihydroxy-2-methyl-propionic acid,etc.in rats.The interconversion of sugar and glucuronic acid,pentose phosphate pathway,fatty acid elongation,fatty acid degradation and tyrosine metabolism play a regulatory role.