QSAR用于功能多肽的结构优化与设计

发布时间：2018-01-25 17:33

本文关键词： 多肽定量构效关系分子设计分子动力学量子化学　出处：《重庆大学》2014年硕士论文　论文类型：学位论文

【摘要】：多肽作为维持生命体正常运转的关键性物质越来越受到人类的关注，特别是其药用价值，并且近年来，其在材料方面的应用价值也日益凸显。但是某些多肽其自身的结构与功能还存在一定的缺陷，因此对其结构的优化与修饰有助于设计出更为合理的多肽化合物。本文运用二维定量构效关系(2D-QSAR)与三维定量构效关系(3D-QSAR)相结合的方法对阿片肽与淀粉样肽两种多肽进行深入研究。主要采用多元线性回归(MLR)、支持向量机(SVM)、Topomer CoMFA及Topomer Search和分子对接等方法对上述两种肽的定量构效关系及作用机理全面分析。并在此基础上，对多肽的结构进行优化，设计出功能更优，结构更为合理的新化合物，结合分子动力学和量子化学，对设计分子进行评估。取得的研究成果如下： ①运用基于R基团的3D Topomer CoMFA建模方法对30个delta阿片六肽进行建模，得到最优模型的拟合、交互验证的复相关系数分别为0.892和0.596。基于所建模型，结合Topomer Search虚拟筛选方法，设计215个非天然氨基酸修饰的delta阿片拟肽。经预测，214个分子的活性高于模板分子。运用Surflex-Dock分子对接研究关键位点作用模式与机制，结合密度泛函理论(DFT)计算相关位点作用能，，结果表明delta阿片肽N末端第三个氨基酸对活性具有重要的贡献。受体的Lys108, Asp128, Tyr129, Lys214, Val217, Val218, Trp284, Leu300, Ile304, Tyr308氨基酸对配体的选择性及活性具有重要的作用。单点能的计算结果显示His278可能与YP型delta阿片肽的活性密切相关。 ②应用FASGAI描述符对30个delta阿片六肽进行序列结构表征，结合逐步回归(SMR)筛选变量，建立MLR模型和偏最小二乘(PLS)模型。经对比，8变量MLR建模结果较优。最优模型的拟合、交互验证的复相关系数分别为0.891和0.756，基于所建模型，设计30条活性较高的delta阿片肽。通过分子对接，研究阿片肽与delta阿片受体可能作用模式。本章从2D的角度进一步证明N末端第三个氨基酸对活性具有重要的影响，5号位和6号位次之。同时运用2D方法设计的分子与运用3D方法设计的分子对受体上作用位点的选择相似。 ③应用FASGAI描述符对277条六肽进行序列结构表征，结合径向基核支持向量机(RBF-SVM)建立淀粉样肽的预测模型，得留五法交互验证的正确率为74.37%，受试者操作特征曲线下面积(AUC)为0.759。对淀粉样肽构效关系分析，氨基酸的疏水性质与静电性质对蛋白质聚集具有关键作用。采用6-残基窗口扫描Aβ42和hIAPP37两种蛋白中的六肽并预测其可聚集性，对比发现，模型识别出可聚集六肽片段与实验报道结果具有较好一致性。基于此模型，设计558条可能聚集的六肽。结合模型预测的Aβ42上的GGVVIA序列和先前设计的8条六肽，运用分子动力学和量子化学方法分析，并辅以圆二色谱(CD)实验验证。结果显示以一种作用方式的能量占优，适度大小的侧链，均匀的能量分布有利于多肽聚集。所设计多肽AEFRHD可能具有较好聚集能力。
[Abstract]:As a key material for maintaining the normal functioning of the living organism , the polypeptide is more and more concerned with the human being , especially its medicinal value , and in recent years , its application in the material has become more and more prominent . However , the structure and function of certain polypeptide have some defects , so the optimization and modification of its structure can help to design a more reasonable polypeptide compound . In this paper , the quantitative structure - effect relationship and mechanism of two kinds of peptides were studied by means of two - dimensional quantitative structure - effect relationship ( 2D - activity relationship ) and three - dimensional quantitative structure - effect relationship ( 3D - quantitative analysis ) . The structure of the polypeptide was optimized by means of multi - linear regression ( MLR ) , support vector machine ( SVM ) , Topomer CoAs and Topomer Search and molecular docking method . The structure of the polypeptide was optimized . The design molecule was evaluated by molecular dynamics and quantum chemistry . The results obtained were as follows : The results showed that the activity of the third amino acid of delta opioid peptide was higher than that of template molecule . The results showed that the third amino acid of delta opioid peptide was higher than that of template molecule . The results showed that the third amino acid of delta opioid N was important to the activity . The results showed that the amino acid of the delta opioid N was highly active in the selection and activity of ligand . The results of single - point energy showed that His278 might be closely related to the activity of delta opioid peptide . ( 2 ) Using the FASGAI descriptor to characterize the sequence structure of 30 delta opioid hexpeptides , combined stepwise regression ( SMR ) screening variables , an MLR model and a partial least squares ( PLS ) model were established . Based on this model , the six peptides of A尾42 and hIAPP37 were analyzed by molecular dynamics and quantum chemistry method .

【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R914.5;R91

【参考文献】