METHODS AND SYSTEMS OF COMPUTATIONAL ANALYSIS FOR PREDICTING CHARACTERISTICS OF COMPOUND
    2.
    发明申请
    METHODS AND SYSTEMS OF COMPUTATIONAL ANALYSIS FOR PREDICTING CHARACTERISTICS OF COMPOUND 审中-公开
    用于预测化合物特性的计算分析方法和系统

    公开(公告)号:US20160146727A1

    公开(公告)日:2016-05-26

    申请号:US14855814

    申请日:2015-09-16

    IPC分类号: G01N21/31 G01N21/33

    摘要: A method for predicting characteristics of a compound includes collecting a first experimental information database for characteristics of reference compounds according to a quantum phenomenon, collecting a simulation database for characteristics of the reference compounds according to the quantum phenomenon by applying density functional theory methods, comparing the simulation database to the first experimental information database for each reference compound to calculate accuracy of the simulation database, clustering the reference compounds into clusters based on the accuracy of the simulation database and designating a proper density functional theory method for each cluster, comparing a similarity between a test compound to predict a characteristic according to the quantum phenomenon and the reference compounds included in each cluster, determining a proper density functional theory method for the test compound according to the similarity, and conducting a simulation with the test compound according to the determined density functional theory method.

    摘要翻译: 一种预测化合物特征的方法包括:根据量子现象收集参考化合物特征的第一实验信息数据库,通过应用密度泛函理论方法,根据量子现象收集参考化合物特征的模拟数据库,比较 模拟数据库为每个参考化合物的第一个实验信息数据库,以计算模拟数据库的准确性,基于模拟数据库的准确性将参考化合物聚类成簇,并为每个簇指定适当的密度泛函理论方法,比较 根据量子现象预测特征的试验化合物和每个簇中包含的参考化合物,根据相似性确定测试化合物的适当的密度泛函理论方法,并用tes进行模拟 t化合物根据确定的密度泛函理论方法。