Strucure Modification And Active Research Of Two Kinds Of Phenolic,QSAR Studies In Phenolic Derivatives In Blumea Ripara Dc

Syringic acid and p-hydroxybenzoic acid are two important components of Guangxi famous Traditional Chinese Medicine "Blumea riparia capsule”.In this paper,syringic acid and p-hydroxybenzoic acid were raw materials.Through structural modification,we hope to get the compounds that have better hemostatic activity.The new compounds YY-A,YY-N,YY-C,YY-D and YY-E were synthesized.The melting points of derivatives were determined,and the structures were characterized by MS,1H-NMR,13C-NMR and X-ray diffraction.From the structure analysis and the adscription,the derivative structures were fully consistent with the theory structure.Some activities of target compounds were tested.In the fluorescence emission spectrum and the synchronous fluorescence spectra,the interaction of HSA with compounds was determined.Following parameters,including quenching mechanism,binging constant,binging sites,binging modes and synchronous fluorescence spectra,were calculated.Plasma recalcification experiment was used to study the plasma interaction of the clove acid derivative.The results showed that YY-A in high concentration showed anticoagulant,low concentration showed coagulation,and YY-N at test concentration maily showed anticoagulant effect.The binding situation between every compounds and HSA was calculated by AutoDock technology.In order to get the relationship between the structure of compounds and the Hemostatic activity,some quantitative structure activity relationships(QSAR)was researched.All synthetic phenolic derivatives were from our group in recent years.First of all,all the compounds were optimized.And surface approx,surface grid,volume,hydration energy,Log P,refractivity,polarizability,mass,lumo,homo,energy were selected as the descriptors.Energy,surface grid and Log P were the most descriptors of APTT,FIB,PT,TT by using Principal Component Regression Analysis,Partial Least Squares Regression and K-Nearest Neighbor Discrimination Method.Through modeling,partial least squares regression equation was obtained.Using leave-one-out method,the hemostatic data were predicted.The effe ct of each descriptor data on hemostasis were got by the Artificial Neural Network Training.