Quantitative Structure-activity Relationship And Molecular Simulation Study Of Enzyme Inhibitors

Computer-aided drug design(CADD) is based on chemistry, life science and computer science, combining with several basic subjects such as mathematics, physics, supplemented by advanced physical and chemical testing technology, which makes CADD become a new systematic subject. CADD computerizes the thinking and methods of rational drug design, providing a strong tool and instrument for the rational drug design. Starting with the fundamental theory of each subject, CADD could provide a very intuitive design line, avoid the blindness of new drug research, save the manpower and material resources, and also accelerate the rate of new drug research and development at the same time. In this paper, quantitative structure-activity relationship and molecular simulation were using as the main method to study the structureactivity relationship about three kinds of enzyme inhibitors. The research of this paper explored the mechanism between drug and related enzyme, and provided theoretic guidance for the optimization of lead compounds. The dissertation includes three main parts as follows: 1: RASMS(random sampling analysis on molecular surface) was used to describe the structure-activity relationship of 65 imidazo[4,5-b]pyridine derivatives as Aurora kinase A inhibitor. Here a QSAR model was built by multiple linear regression(MLR). The estimation stability and prediction ability of the model was strictly analyzed by both internal and external validations. The correlation coefficients of established MLR model, leave-one-out(LOO) crossvalidation and predicted values versus experimental ones of external samples are r2 = 0.782, rCV2 = 0.737 and Qext2 = 0.775, respectively. Three new molecules with higher inhibitory activity were designed by using the QSAR results. Furthermore, satisfactory results show that RASMS could preferably express the information related to the biological activity.2: Molecular docking and 3D-Ho VIAF was used to find the action mode of 36 letrozole derivatives and 34 anastrozole derivatives between aromatase. Here one quantitative structure activity relationship(QSAR) model was by multiple linear regression(MLS).The estimation stability and generalization ability of the model was strictly analyzed by both internal and external validations. For the two groups of compounds, the correlation coefficient of established MLS model(r), leave-one-out cross-validation(rCV), predicted values versus experimental ones of external samples(Qext) were 0.863, 0.782, 0.796 and 0.931, 0.825, 0.641. The results indicated that QSAR model had both favorable estimation stability and good prediction capabilities. By using Auto Dock 4.2 software, we found the best action mode to brug between acceptor. These results revealed that both models have good predictive capability to guide the design and structural modification of homologic compounds. 3: Human African trypanosomiasis(HAT) is a neglected tropical disease, caused by protozoa of species Trypanosoma brucei and spread rapidly on the African continent, but there were no new drugs registered for HAT in the last 20 years. By using Comparative molecular field analysis(Co MFA), and holographic-QSAR(HQSAR), we also developed quantitative structure-activity relationship(QSAR) model for new antiprotozoal benzyl phenyl ether diamidine derivatives. The QSAR model shows not only significant statistical quality, but also satisfies predictive ability: the best Co MFA model had r2=0.915 and rcv2=0.618 and the best HQSAR model had r2=0.951 and rcv2=0.935. Moreover, homology modeling of Trypanosoma brucei type IIα topoisomerase was constructed on the basis of the X-ray crystal structure of the saccharomyces cerevisiae topoisomerase(PDB ID: 4GFH), the most active compound was docked into the homology model. The molecular interactions of the most active compound with the active site residues were discussed in details. Finally, combine with QSAR model and binding mode have shown, 5 new compounds with high activity were successfully designed. Together these results may offer some useful theoretical information in designing potential inhibitors and would be helpful to understand the action mode of diamidine derivatives.