计算机分子模拟技术及人工智能在药物研发中应用.docVIP

计算机分子模拟技术及人工智能在药物研发中应用 　　摘 要：通过计算机模拟手段进行分子对接、药物筛选、先导物的优化、定量构效关系和药效团模型等药物设计方法，可以揭示药物与受体靶标的作用机制，探索药物靶点的空间结构，最终目标是设计具有能选择性地与某一靶标结合的分子；利用分子模拟技术来构造、显示、分析和储存复杂的分子模型，在三维空间中观测药物小分子的结构特征，更改小分子形状和方位，并探测小分子与受大分子靶点的作用机制，判断药物小分子与受体大分子结合的可能活性位点，还能对药物小分子的结构进行修正，提出改善药物的药效学和动力学性质的方案，在“三维空间”中实现直观、可视化的药物分子设计。人工智能利用大数据和机器学习方法，根据已有的药物研发数据自动设计出上百万种与特定靶标相关的小分子化合物，并根据药效、选择性、ADME等其他条件对化合物进行筛选。而后筛选出来的化合物会被合成并且进行实验检测，然后实验数据会被反馈到AI系统中用于改善下一轮化合物的选择。 　　关键词：分子模拟；药物设计；人工智能 　　中图分类号：TP39 文献标志码：A 文章编号：2095-2945（2018）02-0046-02 　　Abstract： Molecular docking， drug screening， optimization of precursors， quantitative structure-activity relationship and pharmacophore model can be used to reveal the mechanism of action between drug and target. To explore the spatial structure of drug targets， the ultimate goal is to design molecules that can selectively bind to a certain target； Using molecular simulation technology to construct， display， analyze and store complex molecular models， observe the structural characteristics of drug small molecules in three-dimensional space， change the shape and orientation of small molecules. The mechanism of the interaction between small molecules and macromolecular targets was explored， the possible active sites of drug small molecules binding to receptor macromolecules were determined， and the structure of drug small molecules could be modified. A scheme to improve the pharmacodynamics and kinetic properties of drugs was put forward to realize the intuitive and visual design of drug molecules in three-dimensional space. Artificial intelligence uses big data and machine learning methods to automatically design millions of small molecular compounds associated with specific targets based on existing drug development data. The compounds were screened according to their pharmacodynamics， selectivity， ADME and other conditions. The selected compounds are then synthesized and tested experimentally， and the experimental data is fed back into the AI system to improve the selec