Suppr超能文献

当不能依赖玻尔兹曼分布时。通过化学位移差异自动进行风信子素的 CASE-3D 结构阐明。

When not to rely on Boltzmann populations. Automated CASE-3D structure elucidation of hyacinthacines through chemical shift differences.

机构信息

Universidade Federal de Pernambuco, Departamento de Química Fundamental, CCEN, Recife, Pernambuco, Brazil, 50670-90.1.

出版信息

Magn Reson Chem. 2020 Feb;58(2):139-144. doi: 10.1002/mrc.4951. Epub 2019 Nov 10.

Abstract

An Akaike Information Criterion (AIC) procedure (CASE-3D) has been successfully applied to the NMR based configurational assignment of reported hyacinthacines (1-3,5-8), recently target of configurational analysis using the popular DP4+ methodology. The present analysis makes use of reported H and C shifts and, in some particular cases, a few J couplings. The difficulty in proper computational prediction of relative energies, in molecules capable of inter-molecular hydrogen bonding, introduces large errors in the prediction of conformationally averaged NMR properties in methods based on Boltzmann averaging such as DP4 or DP4+. In contrast CASE-3D conformational amplitudes are free parameters in the model. Here we show that the CASE-3D conformational model selection strategy, when combined with a larger energy cutoff in the molecular-modelling conformational exploration, was sufficient to correctly assign the relative configuration in five of seven cases. Introduction of more information, either by supplementing H and C data with a few J-couplings, or using a cutoff based on computed DFT energies for the definition of the conformational ensembles, allowed the safe assignment of configuration for all compounds.

摘要

一种 Akaike 信息准则(AIC)程序(CASE-3D)已成功应用于基于 NMR 的报道的杂种碱(1-3、5-8)的构型分配,最近使用流行的 DP4+ 方法对其构型分析成为目标。本分析利用了报道的 H 和 C 位移,在某些特殊情况下,还利用了一些 J 耦合。在能够进行分子间氢键的分子中,正确预测相对能量的计算存在困难,这会导致基于 Boltzmann 平均的 DP4 或 DP4+等方法预测构象平均 NMR 性质的误差较大。相比之下,CASE-3D 的构象幅度是模型中的自由参数。我们在这里表明,当将 CASE-3D 构象模型选择策略与分子建模构象探索中的更大能量截止值相结合时,足以在七种情况下的五种情况下正确分配相对构型。通过补充一些 J 耦合来增加 H 和 C 数据,或者使用基于计算 DFT 能量的截止值来定义构象系综,引入更多信息,可以安全地分配所有化合物的构型。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验