作者: Romano T. Kroemer , Peter Hecht , Klaus R. Liedl
DOI: 10.1002/(SICI)1096-987X(199608)17:11<1296::AID-JCC2>3.0.CO;2-H
关键词: Ab initio quantum chemistry methods 、 Mulliken population analysis 、 Atomic physics 、 Field (physics) 、 Electron 、 MNDO 、 Quantitative structure–activity relationship 、 Chemistry 、 Charge (physics) 、 Ab initio
摘要: Comparative molecular field analysis (CoMFA) is a three-dimensional quantitative structure-activity relationship (3D-QSAR) method which correlates precalculated fields surrounding set of molecules with some target property. Among others, the electrostatic are commonly used. These usually generated by calculating Coulomb potential between probe and bearing atom-centered point charges. The present study was performed in order to investigate up extent different methods calculate potentials can influence results CoMFA. Therefore, variety charge calculation applied data consisting 37 ligands benzodiazepine receptor inverse agonist-antagonist active site. included Gasteiger-Marsili, semiempirical (MNDO, AM1, PM3), ab initio (HF/STO-3G, HF/3-21G*, HF/6-31G*) Semiempirical as well electron populations were derived both from Mulliken population (MPA) or fitting charges (ESPFIT charges). In addition, (MEPs) resulting calculations mapped directly onto CoMFA grid. With regard cross-validated r(2) values (r(2) cv ) QSAR models, ESPFIT-derived yielded generally higher than those MPA For example, at HF/3-21G* level rose 0.61 (MPA-derived potentials) 0.76 fields). MEPs grid not superior corresponding potentials. ESPFIT appeared be similar quality compared CoMFAs. When no scaling steric descriptor matrices applied, contributions influenced high degree magnitude values. © 1996 John Wiley & Sons, Inc.