摘要:流感是经流感病毒引起的一种对人类身体健康方面有众多危害的急性病毒呼吸道传染病，具有传染性强、传播速度快的特点，通常经空气传播。本文运用计算机辅助药物设计的方法，再综合相应的3D-QSAR模型、分子对接和药效团模拟的结果，设计出高效新型流感病毒神经氨酸酶新位点的抑制剂药物。42092

实证研究结果显示，CoMFA模型交叉验证系数q2=0.537，其非交叉验证相关系数R2=0.994，标准偏差为0.116，且数据组间的平均平方误差与数据组内部的平均平方误差的比值F为552.875。CoMSIA模型的交叉验证系数q2=0.516，R2=0.986，标准偏差为0.175， F值为267.094。

毕业论文关键词:3D-QSAR；分子对接；药效团；计算机辅助药物设计

Rational design of Tamiflu derivatives targeting at the open conformation of neuraminidase subtype 1

Abstract：Influenza is caused by a virus of a serious hazard to human health viral respiratory infection,transmitted through the airacutely.In this paper,the method of computer aided drug design,to integrate 3D-QSAR model,molecular docking and pharmacophore group simulation, the result is that we designed a high-efficiency,low toxicity model of statins. Therefore,the 3D-QSAR model is a fast and effective prediction and an essential generic toolsof designingnew drugs.It can provide data for the selection and design of new flu drugs, reducingthe adverse impact ofthe global spread of flu during the day. It provided a theoretical basis and research directions for our greater understanding and designing neuraminidase inhibitors.

CoMFA model cross validation coefficient of q2 =0.537,the non-cross validated coefficient R2=0.994,the standard deviation was 0.116, the ratio between the internal data set of mean square error and data set of mean square error of the F is 552.875. The CoMSIA model of cross validation coefficient q2 =0.516,R2=0.986, the standard deviation was 0.175, F is 267.094. In this experiment,neuraminidase were pided into two groups,namely the open-loop receptor protein and closed-loop receptor protein. With open-loop receptor protein,20 new molecules were designed to determine whether there are 150-cavity. In pharmacophore model,we selected 10 molecules with different activities and structures for virtual screening,final get two positive molecules.

Key Words:3D-QSAR; molecular docking; pharmacophore; computer aided drug design

目录

1. 前言 1

1.1研究背景与意义 1

1.2研究现状 2

1.2.1新型神经氨酸酶抑制剂 3

1.2.2靶向新的结合位点，修饰现有抑制剂的结构 4

1.2.3基于结构的计算机虚拟筛选技术来发现新化学分子 5

1.3研究的基本内容 6

1.3.1数据集的收集 7

1.3.2模型构建的准备工作及分子叠合，

    建立training(训练集)，test(测试集) 7

1.3.3CoMFA模型的建立 7

1.3.4 CoMSIA模型的建立 9

1.3.5训练集预测活性计算 10

1.3.6测试集预测活性计算及新化合物的设计 10

2. 理论原理及计算方法 11

2.1  SYBYL软件 11

2.2  三维定量构效关系 11

2.3 比较分子场分析方法、比较分子相似性指数分析

    与假想活性位点阵方法