Atom(片上系统)
分配系数
集合(抽象数据类型)
分子
分拆(数论)
训练集
计算机科学
化学
数据挖掘
数学
人工智能
组合数学
色谱法
嵌入式系统
有机化学
程序设计语言
作者
Scott A. Wildman,Gordon M. Crippen
出处
期刊:Journal of Chemical Information and Computer Sciences
[American Chemical Society]
日期:1999-08-19
卷期号:39 (5): 868-873
被引量:1266
摘要
We present a new atom type classification system for use in atom-based calculation of partition coefficient (log P) and molar refractivity (MR) designed in part to address published concerns of previous atomic methods. The 68 atomic contributions to log P have been determined by fitting an extensive training set of 9920 molecules, with r2 = 0.918 and σ = 0.677. A separate set of 3412 molecules was used for the determination of contributions to MR with r2 = 0.997 and σ = 1.43. Both calculations are shown to have high predictive ability.
科研通智能强力驱动
Strongly Powered by AbleSci AI