Bame Jessica, Hoeck Casper, Carrington Matthew J, Butts Craig P, Jäger Christof M, Croft Anna K
University of Bristol, School of Chemistry, Clifton, Bristol BS8 1TS, UK.
Phys Chem Chem Phys. 2018 Mar 14;20(11):7523-7531. doi: 10.1039/c7cp07265a.
The use of molecular dynamics (MD) calculations to derive relative populations of conformers is highly sensitive to both timescale and parameterisation of the MD. Where these calculations are coupled with NOE data to determine the dynamics of a molecular system, this can present issues if these populations are thus relied upon. We present an approach that refines the highly accurate PANIC NMR methodology combined with clustering approaches to generate conformers, but without restraining the simulations or considering the relative population distributions generated by MD. Combining this structural sampling with NOE fitting, we demonstrate, for S-adenosylmethionine (aqueous solution at pH 7.0), significant improvements are made to the fit of populations to the experimental data, revealing a strong overall preference for the syn conformation of the adenosyl group relative to the ribose ring, but with less discrimination for the conformation of the ribose ring itself.
使用分子动力学(MD)计算来推导构象异构体的相对丰度对MD的时间尺度和参数化都高度敏感。当这些计算与NOE数据相结合以确定分子系统的动力学时,如果依赖于这些丰度,可能会出现问题。我们提出了一种方法,该方法改进了高度精确的PANIC NMR方法,并结合聚类方法来生成构象异构体,但不限制模拟或考虑MD生成的相对丰度分布。将这种结构采样与NOE拟合相结合,我们证明,对于S-腺苷甲硫氨酸(pH 7.0的水溶液),在将丰度拟合到实验数据方面有显著改进,揭示了相对于核糖环,腺苷基团的顺式构象有很强的总体偏好,但对核糖环本身的构象区分较少。