Nassour Ayoub, Domagala Slawomir, Guillot Benoit, Leduc Theo, Lecomte Claude, Jelsch Christian
Laboratoire de Cristallographie, Résonance Magnétique et Modélisations (CRM2), CNRS, UMR 7036, Institut Jean Barriol, Faculté des Sciences et Technologies, Université de Lorraine, BP 70239, 54506 Vandoeuvre-lès-Nancy Cedex, France.
Department of Chemistry, University of Warsaw, ul. Pasteura 1, 02-093 Warsaw, Poland.
Acta Crystallogr B Struct Sci Cryst Eng Mater. 2017 Aug 1;73(Pt 4):610-625. doi: 10.1107/S2052520617008204. Epub 2017 Jul 27.
A database describing the electron density of common chemical groups using combinations of real and virtual spherical atoms is proposed, as an alternative to the multipolar atom modelling of the molecular charge density. Theoretical structure factors were computed from periodic density functional theory calculations on 38 crystal structures of small molecules and the charge density was subsequently refined using a density model based on real spherical atoms and additional dummy charges on the covalent bonds and on electron lone-pair sites. The electron-density parameters of real and dummy atoms present in a similar chemical environment were averaged on all the molecules studied to build a database of transferable spherical atoms. Compared with the now-popular databases of transferable multipolar parameters, the spherical charge modelling needs fewer parameters to describe the molecular electron density and can be more easily incorporated in molecular modelling software for the computation of electrostatic properties. The construction method of the database is described. In order to analyse to what extent this modelling method can be used to derive meaningful molecular properties, it has been applied to the urea molecule and to biotin/streptavidin, a protein/ligand complex.
提出了一个使用真实和虚拟球形原子组合来描述常见化学基团电子密度的数据库,作为分子电荷密度多极原子建模的替代方法。通过对38个小分子晶体结构进行周期性密度泛函理论计算来计算理论结构因子,随后使用基于真实球形原子以及共价键和电子孤对位点上附加虚拟电荷的密度模型对电荷密度进行精修。对所有研究分子中处于相似化学环境的真实原子和虚拟原子的电子密度参数进行平均,以建立一个可转移球形原子数据库。与目前流行的可转移多极参数数据库相比,球形电荷建模需要更少的参数来描述分子电子密度,并且可以更轻松地纳入分子建模软件中以计算静电性质。描述了该数据库的构建方法。为了分析这种建模方法在多大程度上可用于推导有意义的分子性质,已将其应用于尿素分子以及生物素/链霉亲和素(一种蛋白质/配体复合物)。