Accurate energy evaluation at atomic level and establishing NMR chemical shifts as reliable reporters of atomic charge distributions

该项目有两个目标。首先是开发下一代力场，使其能够在原子水平上进行准确的能量评估。这一目标将与工业合作伙伴阿斯利康（见下一节）进行探讨。第二是使用相关的方法来确定蛋白质中的核磁共振测量如何以及在多大程度上可以用作原子水平上电荷分布的可靠替代报告。对于第二个目标，我们将专注于在酶内提供催化的分子相互作用。虽然酶催化的知识在不断扩大，但与分子内和分子间相互作用的量化有关的基本问题仍然限制了预测其催化行为的能力。酶活性位点上电荷的分布，它与原子位置的关系，以及它在催化循环中如何变化，对于建立一定程度的理解，从而可靠地设计新的催化剂至关重要。目前，为了准确地了解和预测催化速率的增强，需要将非常高分辨率的结构生物学和非常高水平的量子计算结合起来，并使用精确的方法将计算出的电子密度划分到单个原子。所需的分辨率仅在实验确定的蛋白质结构中很少可用，并且当前的计算能力表明，在适当的量子计算中可以容纳相对较少的原子（<200），这意味着生成的图像必然是有限的。我们的初步工作已经确定，使用量子化学拓扑（QCT）方法衍生的分子原子量子理论（QTAIM）方法可以提供定量可预测框架的证据，将核磁共振化学位移与原子电荷联系起来。现在的重点是利用这些令人兴奋的初步数据，并将QTAIM推导的预测扩展到酶，在那里我们有必要的实验NMR和x射线数据，可以用于代表催化循环的各种复合物，以及适当的QM模型。在催化循环的不同阶段，对原子电荷分布的理解，以及它们与分子结构和键形成/裂解的关系，对催化的未来开发具有深远的影响。回答这些问题的方法（学生实际会做什么）；REG方法是由Popelier小组开发的，并附带了一个名为REG.py的内部代码。学生将使用Python编程，修改内部代码REG.py，使其与晶体结构预测问题相连接。他将研究REG过程的简化和加速，从而大大减少计算时间。学生将参与仔细和系统的测试，从而获得前所未有的量子力学洞察力。研究的新颖工程和/或物理科学内容（将其置于EPSRC职权范围内的科学）。该项目隶属于化学科学大挑战“定向组装具有目标特性的扩展结构（DAESTP）”。该项目的第一个目标有很强的机器学习成分，因此与人工智能重叠，这是一个热门的融资主题。相关的科学产品被称为FFLUX，这是一个全新的力场，由新颖的原理设计并编码为软件包。

The project has two objectives. The first is to develop a next generation force field that will allow accurate energy evaluation at atomic level. This objective will be explored with the industrial partner AstraZeneca (See next section). The second is to use related approaches to establish how and to what extent NMR measurements in proteins can be used as a reliable surrogate reporter of the charge distributions at atomic level.For the second objective, we will focus on molecular interactions that deliver catalysis within enzymes. While knowledge of enzyme catalysis is ever-expanding, fundamental questions relating to the quantification of intra- and intermolecular interactions still limit the ability to predict their catalytic behaviour. The distribution of charge in enzyme active sites, how it relates to atomic positions, and how it changes during the catalytic cycle, are of critical importance to develop a level of understanding where new catalysts can be reliably designed. Currently, access to such information at the accuracy required to understand and predict catalytic rate enhancements dependably, requires a combination of very high resolution structural biology and very high level quantum calculations, along with accurate methods to partition the calculated electron densities to individual atoms. The required resolution is only seldomly available in experimentally determined protein structures, and current computational capabilities dictate that relatively few atoms (<200) can be accommodated in the appropriate quantum calculations, meaning that the generated picture is necessarily limited.Our preliminary work has established that charge partitioning using the Quantum Theory of Atoms in Molecules (QTAIM) method, derived from the Quantum Chemical Topology (QCT) approach can deliver proof of a quantitative predictable framework linking NMR chemical shifts to atomic charges. The focus is now to take these exciting preliminary data and extend QTAIM derived predictions to enzymes where we have the necessary experimental NMR and X-ray data available for various complexes that represent the catalytic cycle, and appropriate QM models for these. The derived understanding of atomic charge distributions, and how they relate to molecular structure and bond formation / cleavage, at different stages of the catalytic cycle has far reaching consequences for the future exploitation of catalysis.the approach that will be taken to answer these questions (what the student will actually be doing); The REG method has been developed in the Popelier group and comes with an in-house code called REG.py. The student will be programming in Python, modifying the in-house code REG.py to interface them with the problem of crystal structure prediction. He will investigate the streamlining and speeding up of the REG procedure leading to a severe reduction of computation time. The student will be involved with careful and systematic testing and thus gather unprecedented quantum mechanical insight.the novel engineering and/or physical sciences content of the research (the science that places it within EPSRC's remit).This project resorts under the Chemical Sciences Grand Challenge of "Directed Assembly of Extended Structures with Targeted Properties (DAESTP)". There is a strong Machine Learning component to the first objective of the project, and thus overlap with Artificial Intelligence, a popular funding topic. The associated scientific product is called FFLUX, which is a completely new force field, designed by novel principles and encoded as a software package.