Ermanis K, Parkes K E B, Agback T, Goodman J M
Centre for Molecular Science Informatics, Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, UK.
Org Biomol Chem. 2017 Oct 31;15(42):8998-9007. doi: 10.1039/c7ob01379e.
A large-scale optimisation of density functional theory (DFT) conditions for computational NMR structure elucidation has been conducted by systematically screening the DFT functionals and statistical models. The extended PyDP4 workflow was tested on a diverse and challenging set of 42 biologically active and stereochemically rich compounds, including highly flexible molecules. MMFF/mPW1PW91/M06-2X in combination with a 2 Gaussian, 1 region statistical model was capable of identifying the correct diastereomer among up to an upper limit of 32 potential diastereomers. Overall a 2-fold reduction in structural uncertainty and a 7-fold reduction in model overconfidence have been achieved. Tools for rapid set-up and analysis of computational and experimental results, as well as for the statistical model generation, have been developed and are provided. All of this should facilitate rapid and reliable computational NMR structure elucidation, which has become a valuable tool to natural product chemists and synthetic chemists alike.
通过系统筛选密度泛函理论（DFT）泛函和统计模型，对用于计算核磁共振（NMR）结构解析的DFT条件进行了大规模优化。扩展的PyDP4工作流程在一组多样且具有挑战性的42种生物活性和立体化学丰富的化合物上进行了测试，包括高度灵活的分子。MMFF/mPW1PW91/M06-2X与双高斯、单区域统计模型相结合，能够在多达32种潜在非对映异构体的上限范围内识别出正确的非对映异构体。总体而言，结构不确定性降低了2倍，模型过度自信程度降低了7倍。已经开发并提供了用于快速设置和分析计算及实验结果以及生成统计模型的工具。所有这些都应有助于快速可靠地进行计算NMR结构解析，这已成为天然产物化学家和合成化学家都非常有价值的工具。
