X-ray data analysis in the presence of structural variability

 说明(申请人提供)：“存在结构可变性的X射线数据分析”提案旨在推进衍射数据分析方法，以便在倒易空间的数据处理过程中和在真实空间的结构分析过程中对晶体之间和晶体内的可变性进行最佳建模。拟议工作的重要性源于该技术的重要性，该技术可生成独特的详细信息。X射线结构被用来直接在原子水平上了解细胞过程，解释和验证通过其他技术获得的结果，为细胞过程的详细研究产生假设，并指导药物设计研究--所有这些都与NIH的任务高度相关。大分子晶体通常具有有限的尺寸和晶格有序性。两者都可能导致需要组合来自多个晶体的数据以获得成功的结构解，有限的级数产生衍射伪影，并与不同样品之间的非同构相关。非同构阻碍了对来自多个晶体的数据集进行平均，因为为了成功地进行平均，数据需要非常相似。由于单一数据集中数据的不完整性、被研究晶体中辐射引起的变化，以及缺乏统计方法来告知实验人员数据分析是否在正确的方向上进行，平均化的问题变得更加复杂。当需要分析大量数据集时，由于数据分析的组合复杂性，还存在与平均多个数据集相关的技术挑战。最后的困难出现在分析从多晶体实验获得的结构结果时，必须将所需的生物信号从噪声中分离出来，例如配体的存在或分子的特定动力学行为。我们的提案通过开发和实施创新方法来解决这些问题。在目标1中，将开发和实施新的方法，对一个或多个晶体产生的多个可能不完整的数据集进行平均。由于我们对非同构的组件进行建模的创新方法，我们预计即使是完全不同构的数据集也可以一起用于解决具有挑战性的结构。在目标2中，将开发分析真实空间中多个晶体的平均数据集的结果的方法。平均的描述符将与结构分析的结果相关联，从而可以量化和解释造成真实空间可变性的因素。最后，在AM 3中，将开发一个基于Web的服务器，以便向结构生物学社区提供这些方法。

 DESCRIPTION (provided by applicant): The proposal "X-ray data analysis in the presence of structural variability" aims to advance diffraction data analysis methods so that the variability between crystals and within crystals is optimally modeled during data processing in reciprocal space and during structural analysis in real space. The significance of the proposed work results from the importance of the technique, which generates uniquely-detailed information. X-ray structures are used to understand cellular processes at the atomic level directly, to explain and validate results obtain by other techniques, to generate hypotheses for detailed studies of cellular process, and to guide drug design studies - all of which are highly relevant to the NIH mission. Macromolecular crystals are frequently of limited size and crystal lattice order. Both may result in the need for combining data from multiple crystals for successful structure solution, with the limited order generating diffraction artifacts and correlating with non-isomorphism between different specimens. Non-isomorphism hinders the averaging of data sets from multiple crystals, because for successful averaging, data need to be very similar. The problems with averaging are compounded by incompleteness of the data in a single data set, radiation-induced changes in the crystal under investigation, and lack of statistical measures that would inform experimenters regarding whether or not the data analysis is progressing in the right direction. There are also technical challenges associated with averaging multiple data sets that result from the combinatorial complexity of data analysis when a large number of data sets need to be analyzed. Final difficulty appears when the analysis of the structural results obtained from multi-crystal experiments must separate the desired biological signals, e.g. the presence of a ligand or a specific dynamic behavior of the molecules, from the noise. Our proposal addresses these problems by developing and implementing innovative approaches. In Aim 1, new approaches will be developed and implemented for averaging multiple, potentially incomplete data sets resulting from one or more crystals. Owing to our innovative approach to modeling the components of non-isomorphism, we expect that even quite non-isomorphous data sets can be used together to solve challenging structures. In Aim 2, methods that will analyze the results of averaging data sets from multiple crystals in real space will be developed. The descriptors of averaging will be correlated with the outcomes of the structural analysis, so that the contributors to variability in real space can be quantified and interpreted. Finally, in Am 3, a web-based server will be developed in order to provide these methods to the structural biology community.