Studies Of Quantitative Structure-activity Relationship Of Hepatitis C Polymerase And Checkpoint Kinase1Inhibitiors

The study of quantitative structure-activity relationship (QSAR) is one of the hotissues in chemometrics and bioinformatics research areas. It is widely used in variousareas, such as biochemistry, environmental science, analytical chemistry and so on.Especially, its application is more prominent in the field of computer-aided drugdesign. The main purpose of QSAR is to study the relationship of molecular structureand their properties (biologicalactivity, physical properties, chemical properties, ect.)of the target compounds by using different chemometrics methods. In this thesis, wehave discussed the relationship between structures and biological activities ofpolymerase and protein kinase inhibitors, respectively. And the different methodswere adopted to construct QSAR models. What's more, the QSAR models werequickly, accurately to predict the activities of unknown compounds. As well as, themain structural elements affecting activities of inhibitors were discussed. Therefore, itwill provide the new information to design and synthesize higher activity inhibitors.The thesis was composed of four chapters. The first chapter reviewed theresearch progress and application of quantitative structure-activity relationship(QSAR) in the multidisciplinary. In addition, it also described the fundamentals ofQSAR. What's more, the procedures and approaches of the construction of QSARmodels were described in details.The second chapter studied the relationship between molecular structures andinhibitory activity of52pyridazinones derivatives inhibitors based on HCV NS5Bpolymerase. In this chapter, stepwise multiple linear regression (stepwise-MLR)method and uninformative variable elimination-partial least squares (UVE-PLS) wereused to select the characteristic parameters of compounds, respectively. Subsequently,partial least squares (PLS) and particle swarm optimization support vector machine(PSO-SVM) were utilized to construct the linear and nonlinear models by two set ofthe selected descriptors and their activity data, respectively. The predictiveperformance of the proposed models was evaluated by the strict criteria. Finally, thepredictive power of the PSO-SVM models was better than the corresponding PLSmodels. Its object was to construct the robust and reliable QSAR models and predict the activity of the unknown pyridazinone derivatives. As well as, we also discussedthe main structural factors which are associated with the inhibitor activity. Namely,inhibitory activity of pyridazinone derivatives was closely related to electronegativity,molecular mass, atomic polarizability, atomic van der Waals volumes and3Dinformation of molecular structures.In the third chapter, the relationship between molecular structures and inhibitoryactivity of36pyrazolo [1,5-a] pyrimidine derivative inhibitors of checkpoint kinase1(Chk1) was investigated by using QSAR approaches. Based on the Kennard-Stone(KS) algorithm, the entire data set which was composed of36compounds wasdivided into training set (24compounds) and test set (12compounds). Partial leastsquares (PLS) and particle swarm optimization support vector machine (PSO-SVM)were used to construct QSAR models via the molecular descriptors and half inhibitoryconcentration (IC50) of pyrazole [1,5-a] pyrimidine derivatives. Finally, comparedwith PLS, the performance of the PSO-SVM model was better. Thus, it can beinferred that the PSO-SVM analysis will be a promising method and be extensivelyapplied into the QSAR studies.In the last work, a series of triazolone derivatives as novel checkpoint kinase1inhibitors were regarded as research subjects in QSAR studies. Three-dimensionalQSAR (3D-QSAR) model was constructed by the relationship between descriptorsand biological activity of compounds. What's more, CoMFA approach was used toexplain the impact of molecular field on inhibitory activity of triazolone derivatives.As well as, we find out the key structural factors of compounds are closely related toinhibitory activity. Hence, a reliable theoretical basis is proposed to design andsynthesize a novel, highly active checkpoint kinase1inhibitor.