基于分子指纹的化学结构相似度检索系统的研究(英文)

Research of chemical similarity searching system based on molecular fingerprint

化学结构相似性检索在现代化学研究中具有重要作用。而化学结构的相似性度量是进行相似度检索的基础和前提。目前在化学信息学的研究中,有数量众多的化学结构距离度量和相似度表示方法。本文采用了Daylight的分子指纹方法,并采用了Tanimoto系数定义的相似度度量方法。并根据这种度量方法使用CDK来进行计算化学结构的相似度指数。在上述研究工作的基础上,开发了基于浏览器/服务器模式的化学结构相似度检索系统,通过该系统可以在中药活性成分数据库中进行化学结构相似度检索。用户在进行化学结构相似度检索时,可以选择已存在的化学结构,也可以采用JME来绘制新的化学结构。下一步将在该数据库中进行聚类分析和分子多样性的研究。

Chemical similarity searching plays an important role in modem chemical research. The similarity measure is fundamental for chemical similarity searching. There are many distance metrics and similarity coefficients commonly used in chemical information systems. In this paper, Daylight fingerprint was adopted as chemical structural representation method. And similarity indexes were calculated according to Tanimoto coefficient definition using the famous chemical library project-Chemical Development Kit （CDK）. Based on these work, chemical similarity searching system was developed in the mode of Browser/Server. In this system, user can fulfill chemical similarity searching by choosing an existing structure or drawing a new structure by JME. There are many applications of the similarity measures and searching, including database clustering, docking searches, reaction similarity searching, and the analysis of molecular diversity. Research of clustering analysis and molecular diversity will be the next work based on the Traditional Chinese Medicine Database （TCMD）.