Resolution of ab initio shapes determined from small-angle scattering.
Resolution of ab initio shapes determined from small-angle scattering.
复制标题
DOI：
10.1107/s2052252516016018
复制
发表时间：
2016-11-01
期刊：
影响因子：
3.9
通讯作者：
Svergun DI
中科院分区：
文献类型：
--
作者：
Tuukkanen AT;Kleywegt GJ;Svergun DI
基于多次重建的可变性，提出了一种小角散射数据从头开始重建形状的定量分辨率度量方法。该方法在仿真算例中得到了验证，并在实验数据中得到了应用。空间分辨率是结构模型的一个重要特征，用x射线晶体学或电子冷冻显微镜确定结构的作者总是在发表和沉积时提供分辨率。x射线或中子的小角散射（SAS）最近成为一种主流的结构方法，可以提供溶液中蛋白质、核酸和复合物的整体三维结构。然而，sas衍生模型没有定量的分辨率测量，这极大地阻碍了它们的验证和进一步使用。本文提出了一种基于散射数据的从头算形状重建分辨率评估方法。利用从头算形状的固有可变性，并演示了它们的平均傅立叶壳相关函数如何与模型分辨率相关。通过对已知高分辨率结构蛋白质的模拟数据验证了该方法的有效性，并在实验数据中得到了应用。建议今后在出版物和从头算SAS模型的沉积中报告该决议。
A quantitative resolution measure of the ab initio shapes restored from small-angle scattering data is introduced based on the variability of multiple reconstructions. The new measure is validated in simulated examples and its efficiency has been demonstrated in applications to experimental data. Spatial resolution is an important characteristic of structural models, and the authors of structures determined by X-ray crystallography or electron cryo-microscopy always provide the resolution upon publication and deposition. Small-angle scattering of X-rays or neutrons (SAS) has recently become a mainstream structural method providing the overall three-dimensional structures of proteins, nucleic acids and complexes in solution. However, no quantitative resolution measure is available for SAS-derived models, which significantly hampers their validation and further use. Here, a method is derived for resolution assessment for ab initio shape reconstruction from scattering data. The inherent variability of the ab initio shapes is utilized and it is demonstrated how their average Fourier shell correlation function is related to the model resolution. The method is validated against simulated data for proteins with known high-resolution structures and its efficiency is demonstrated in applications to experimental data. It is proposed that henceforth the resolution be reported in publications and depositions of ab initio SAS models.