化学学报 ›› 2004, Vol. 62 ›› Issue (17): 1585-1589. 上一篇    下一篇

研究论文

R-基团相似性的比较

朱倩刚, 姚建华刚, 李丰刚, 陈海峰刚, 袁身刚   

  1. 中国科学院上海有机化学研究所计算机化学实验室, 上海, 200032
  • 投稿日期:2004-02-09 修回日期:2004-04-20 发布日期:2014-02-17
  • 通讯作者: yaojh@mail.sioc.ac.cn
  • 作者简介:姚建华E-mail:yaojh@mail.sioc.ac.cn;Tel:+86-021-64163300-2888;Fax:+86-21-64166128.
  • 基金资助:
    本工作由2003CB114401,2002AA231011,02DJ14013,CNRS/CAS14916和"基于现代理论和技术的复方中药系统研究"五个项目所组成,它们分别得到中国科学技术部,上海市科学技术委员会和中国科学院的资助.

Comparison of the R-Group Similarity

ZHU Qian, YAO Jian-Hua, LI Feng, CHEN Hai-Feng, YUAN Shen-Gang   

  1. Laboratory of Computer Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 200032
  • Received:2004-02-09 Revised:2004-04-20 Published:2014-02-17

介绍了对化学取代基(R-基团)进行相似性比较的工作.每个R-基团主要依赖一些对反应影响较为明显,同时又便于计算的结构描述符来进行描述,例如电负性、氢键受体、氢键给体等参数.由于与反应核心相距过远的环境对反应的影响较小,R-基团相似性比较对象,就由每个R-基团截取从它与母体连接的位点出发向外扩展6层的子结构组成.在确立了对R-基团描述和距离限制的基础上,提出了用一个45维向量表示一个R-基团,并由化学结构的比较,转化为向量比较来实现R-基团相似性比较的方法.采用这一方法,成功地对大量的R-基团进行了相似性的区分,并实现了对未知化合物进行相似性预测的目标.

关键词: 相似性, 描述符, 归一化, 重心

This work is mainly concerned with the comparison of the similarities of the R-groups.Each R-group is based on several descriptors, such as electronegativity, hydrogen-bond acceptor, hydrogen-bond donor etc., which will affect reactions obviously and can be calculated conveniently.Since the environment with very long distance from the reaction core has relatively small effect on the reaction, the comparison partners of R-group can be regarded as some substructures that have through-bond of up to six from the point of substitution on the reaction core.This work proposed a method based on the description of R-group and the constraints of distance to compare the R-group similarities, which represented an R-group with a 45 dimensional vector and converted the comparison of chemical structures into the corresponding multiple vectors.Employing this method, we successfully distinguished a large number of R-group similarities, and realized the goal of predicating the similarity of unknown chemical compounds.

Key words: similarity, descriptor, normalization, barycenter