药效团
虚拟筛选
数量结构-活动关系
化学
鉴定(生物学)
计算生物学
计算机科学
人工智能
组合化学
立体化学
生物
植物
作者
Peng Lü,Yubin Wang,Ping Ouyang,Jin-Xiong She,Mingfang He
出处
期刊:Current Computer - Aided Drug Design
[Bentham Science Publishers]
日期:2015-07-22
卷期号:11 (1): 51-56
被引量:7
标识
DOI:10.2174/1573409911666150529125814
摘要
Pharmacophore models of G protein-coupled receptor40 (GPR40) agonists were developed using Discovery Studio V2.1. One hydrogen bond acceptor and three hydrophobic features, Hypo 1 which was the best hypothesis, had a correlation co-efficient of 0.971, cost difference of 73.041, and RMSD 0.680. This model was validated by test set, Fischer randomization test and decoy set. Subsequently, Hypo 1 was employed as a 3D query to identify potent molecules from chembridge database. 21 compounds were identified with estimated EC50 less than 500 nM. Seven top-scored hit compounds were chosen for further evaluation in FLIPR assay and two compounds were discovered as potent GPR40 agonists.
科研通智能强力驱动
Strongly Powered by AbleSci AI