• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文学位 >Invariant representation and perception of molecular surface shape: Applications for ligand docking and similarity searching.
【24h】

Invariant representation and perception of molecular surface shape: Applications for ligand docking and similarity searching.

机译：分子表面形状的不变表示和感知：配体对接和相似性搜索的应用。

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Molecular shape comparison is a fundamental area of research in computer-aided drug design. Applications of molecular shape analysis are numerous and include three-dimensional molecular similarity searching, the docking of a ligand into a receptor protein, and shape-base molecular structure indexing. Current methods of performing molecular shape analysis are either very demanding of computational resources or provide an inadequate representation of molecular shape. The inadequate representation of shape frustrates attempts to develop efficient algorithms for molecular shape analysis. This present work describes a means of efficiently and accurately representing local molecular shape.; We have developed a means of describing local molecular surface shape that is particularly well suited for geometrically-based molecular shape comparison algorithms. The representation of shape presented in this research is based upon approximating small regions of the molecular surface of a molecule by quadratic functions. Geometrically invariant properties of the surface are derived by calculating the differential geometric properties of the quadratic approximations. These properties are then used to construct a quadratic shape descriptor (QSD) that describes local molecular surface shape. Not only is the representation of shape geometrically invariant, but it also provides an efficient mechanism for quantifying the degree of shape similarity between two sections of surface. These properties make QSDs ideally suited for use in molecular shape comparison algorithms.; Using the QSD method of shape representation, we have developed a series of extremely efficient molecular similarity and docking algorithms. The presented algorithms are deterministic and have a low order of computational complexity in both time and space. The similarity algorithms are able to generate experimentally verified three-dimensional alignments of molecules. The rigid and flexible ligand docking algorithms produce crystallographically verified orientations of ligands in their receptor sites. We also demonstrate the use of QSDs for the perception of high level molecular shape.
机译：分子形状比较是计算机辅助药物设计研究的基础领域。分子形状分析的应用很多，包括三维分子相似性搜索，配体对接到受体蛋白中以及基于形状的分子结构索引。当前执行分子形状分析的方法要么对计算资源的要求非常高，要么提供的分子形状表示不充分。形状表示的不足使尝试开发用于分子形状分析的有效算法成为可能。本工作描述了一种有效而准确地表示局部分子形状的方法。我们已经开发出一种描述局部分子表面形状的方法，该方法特别适合基于几何的分子形状比较算法。在这项研究中呈现的形状表示是基于二次函数逼近一个分子的分子表面的小区域。通过计算二次近似的微分几何特性，可以得出表面的几何不变特性。然后将这些属性用于构造描述局部分子表面形状的二次形描述符（QSD）。形状的表示不仅在几何上是不变的，而且还提供了一种有效的机制来量化两个表面部分之间的形状相似度。这些特性使QSD非常适合用于分子形状比较算法。使用QSD的形状表示方法，我们开发了一系列极为有效的分子相似性和对接算法。提出的算法是确定性的，并且在时间和空间上的计算复杂度较低。相似性算法能够生成经过实验验证的分子三维比对。刚性和柔性配体对接算法可在受体位置产生经晶体学验证的配体取向。我们还演示了使用QSD感知高级分子形状。

著录项

  • 作者

    Goldman, Brian Brunetti.;

  • 作者单位

    University of California, Santa Cruz.;

  • 授予单位 University of California, Santa Cruz.;
  • 学科 Chemistry Pharmaceutical.
  • 学位 Ph.D.
  • 年度 1997
  • 页码 243 p.
  • 总页数 243
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类 药物化学;
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号