SI2-CHE: CCP-SAS - Collaborative Computational Project for advanced analyses of structural data in chemical biology and soft condensed matter

在过去的二十年里，美国和英国在高亮度多用户X射线同步加速器和中子源方面的主要基础设施投资取得了巨大的成功，使外部（大学和商业）用户能够利用这些设施在更具挑战性和重要性的系统上进行数据收集。特别重要的是大分子晶体学（MX）领域，其中硬件投资（例如：仅英国钻石光源就有六个MX站）也受益于大规模的集成软件开发，维护和分销，主要是通过CCP4计划。结果是新的晶体结构数据激增。现在需要同样的革命来推进化学生物学和软凝聚态研究，因为我们周围的世界不是静态结构。为了实现新工程和材料科学应用的仪器投资的全部效益，需要将大量的X射线和中子数据与用户友好的，高通量的数据分子建模相结合，以揭示这些系统中的结构如何随着不同的实验条件在时间和空间上发生变化。这是一个问题，定义了大挑战，通过这个建议来解决。合作的动机是公认的重要性，在美国和英国的大型多用户同步加速器和中子设施的结构建模。在生物大分子之外，软凝聚态科学传统上使用相对简单的模型系统。然而，近年来，随着研究人员努力生产可销售的材料，胶体和聚合物科学的复杂性迅速增加，引发了对新的综合分子建模程序的迫切需求，以解释实验数据集。因为这些计算分析通常是由个人在他们的外部机构进行的，所以解释的进展，毫不奇怪，比大分子晶体学中看到的要慢得多。在这里，通过将多用户设施的计算专业知识与需要模拟不同数据集的经验丰富的外部实验室联系起来，我们处于开发必要程序和基础设施的理想位置。这项建议的智力价值在于开发新的算法，以便更准确和快速地分析结构数据，以及软件基础设施，以便能够快速和全面地对从这些学科包括小角X射线和中子散射（SAXS和SANS）、广角散射、分析超浓缩（AUC）和NMR光谱学。事实上，不同实验方法的结合提供了单独使用一种方法无法获得的新见解。高端计算分子建模和模拟的重大进展为创建建模软件基础设施提供了一个令人兴奋的机会，该软件基础设施可以通过不同技术的组合访问令人生畏的信息内容。美国和英国团队汇集了胶体科学家，生物工程师和计算化学家，以及SAXS，SANS，AUC，NMR和高性能计算（HPC）及其计算基础设施的专家，以应对大分子和超分子化学的挑战。我们的多方面方法的更广泛的影响将来自高端分子建模的应用，使用来自各种原位结构探针的数据作为约束。目标是提供软件环境，使多学科实验团队能够详细了解复杂的化学相互作用以及它们如何塑造结构组织。如果成功的话，这有可能改变软物质科学的研究方式，并大大加快这些重要领域的发现过程。

The major infrastructural investment by the US and UK in high brilliance multiuser X-ray synchrotrons and neutron sources during the last two decades has been immensely successful in allowing external (university and commercial) users to exploit these facilities for data collection on ever more challenging and important systems. Of particular significance is the field of macromolecular crystallography (MX) where the hardware investment (eg: the UK Diamond Light Source alone has six MX stations) has also benefitted from large-scale, integrated, software development, maintenance and distribution largely through the CCP4 initiative. The result has been an explosion of new crystal structure data. The equivalent revolution is now needed to advance chemical biology and soft condensed matter research because the world around us is not one of static structures. To realize the full benefit of the instrumental investment for novel engineering and material science applications requires integrating the plethora of X-ray and neutron data with user-friendly, high-throughput, molecular modelling of the data, in order to reveal how structure in these systems changes in time and space with varying experimental conditions. This is the issue that defines the Grand Challenge to be tackled through this proposal.The collaboration is motivated by the recognized importance of structural modelling at the large multiuser synchrotron and neutron facilities in both the US and the UK. Outside of biological macromolecules, soft condensed matter science has traditionally worked with relatively simple model systems. However, in recent years there has been a rapid increase in the complexity of colloid and polymer science as researchers strive to produce marketable materials, triggering a critical need for new integrated molecular modelling procedures to explain experimental data sets. Because these computational analyses are typically performed by individuals at their external institutions, progress in interpretation has, unsurprisingly, been much slower than witnessed in macromolecular crystallography. Here, by linking computational expertise at the multiuser facilities with experienced external laboratories that need to model diverse data sets, we are ideally positioned to develop the necessary procedures and infrastructure.The intellectual merit of this proposal lies in the development of both new algorithms for more accurately and rapidly analysing structural data and the software infrastructure that will enable rapid and encompassing modelling of data obtained from complementary disciplines such as small angle X-ray and neutron scattering (SAXS and SANS), wide angle scattering, analytical ultracentrifugation (AUC) and NMR spectroscopy. Indeed the combination of different experimental methods provides new insights not available from one method alone. The significant advances in high end computational molecular modelling and simulation provide an exciting opportunity to create a software infrastructure for modelling which can make accessible the daunting information content from the combination of different techniques. The US and UK teams bring together colloid scientists, bioengineers and computational chemists, with experts in SAXS, SANS, AUC, NMR, and high performance computing (HPC) and its computing infrastructure for this challenge in macromolecular and supramolecular chemistry. The broader impact of our multifaceted approach will result from the application of high-end molecular modelling using the data from the various in-situ structural probes as constraints. The goal will be to provide the software environment to enable multidisciplinary experimental teams to gain a detailed understanding of complex chemical interactions and how they shape structural organization. This has the potential, if successful, to transform the way soft matter science is done, and to dramatically accelerate the discovery process in these important fields.