Petoukhov Maxim V, Svergun Dmitri I
European Molecular Biology Laboratory, Hamburg Unit, EMBL c/o DESY, Notkestrasse 85, 22603 Hamburg, Germany.
Acta Crystallogr D Biol Crystallogr. 2015 May;71(Pt 5):1051-8. doi: 10.1107/S1399004715002576. Epub 2015 Apr 24.
A novel approach is presented for an a priori assessment of the ambiguity associated with spherically averaged single-particle scattering. The approach is of broad interest to the structural biology community, allowing the rapid and model-independent assessment of the inherent non-uniqueness of three-dimensional shape reconstruction from scattering experiments on solutions of biological macromolecules. One-dimensional scattering curves recorded from monodisperse systems are nowadays routinely utilized to generate low-resolution particle shapes, but the potential ambiguity of such reconstructions remains a major issue. At present, the (non)uniqueness can only be assessed by a posteriori comparison and averaging of repetitive Monte Carlo-based shape-determination runs. The new a priori ambiguity measure is based on the number of distinct shape categories compatible with a given data set. For this purpose, a comprehensive library of over 14,000 shape topologies has been generated containing up to seven beads closely packed on a hexagonal grid. The computed scattering curves rescaled to keep only the shape topology rather than the overall size information provide a `scattering map' of this set of shapes. For a given scattering data set, one rapidly obtains the number of neighbours in the map and the associated shape topologies such that in addition to providing a quantitative ambiguity measure the algorithm may also serve as an alternative shape-analysis tool. The approach has been validated in model calculations on geometrical bodies and its usefulness is further demonstrated on a number of experimental X-ray scattering data sets from proteins in solution. A quantitative ambiguity score (a-score) is introduced to provide immediate and convenient guidance to the user on the uniqueness of the ab initio shape reconstruction from the given data set.
本文提出了一种全新的方法，用于对与球平均单粒子散射相关的模糊性进行先验评估。该方法引起了结构生物学界的广泛关注，它能够对生物大分子溶液散射实验中三维形状重建固有的非唯一性进行快速且与模型无关的评估。如今，从单分散系统记录的一维散射曲线常被用于生成低分辨率的粒子形状，但此类重建的潜在模糊性仍是一个主要问题。目前，（非）唯一性只能通过基于重复蒙特卡罗形状确定运行的后验比较和平均来评估。新的先验模糊性度量基于与给定数据集兼容的不同形状类别的数量。为此，已生成了一个包含超过14000种形状拓扑结构的综合库，其中包含紧密堆积在六边形网格上的多达七个珠子。重新缩放后的计算散射曲线仅保留形状拓扑结构而非整体尺寸信息，从而提供了这组形状的“散射图”。对于给定的散射数据集，可以快速在图中获得相邻形状的数量以及相关的形状拓扑结构，这样该算法除了提供定量的模糊性度量外，还可作为一种替代的形状分析工具。该方法已在几何体的模型计算中得到验证，并且在一些来自溶液中蛋白质的实验X射线散射数据集上进一步证明了其有效性。引入了定量模糊性分数（a分数），以便就从给定数据集进行从头形状重建的唯一性向用户提供即时且便捷的指导。
