A new computational strategy for the modelling of self-assembly biomolecules

金属离子在许多生物和化学过程中起着至关重要的作用。离子不仅是化学研究的基础，在生物分子的组装和功能中也起着至关重要的作用。例子包括蛋白质笼、1、2多酶复合物、3和RNA。此外，在与有机分子结合时，它们还可以作为合成系统组装的基石。金属介导的自组装已经成为一种有前途的方法，可以从更简单的构建块开发复杂的仿生系统。然而，即使是自然发生的自组装，理解驱动自组织、结构和进化的基本规则仍然是一个挑战，这极大地限制了我们重新利用这些系统属性的能力。这个项目的总体目标是。这一目标将通过处理以下几个方面来实现：金属中心精确力场的开发（1-9个月）。2. 实施有效的建模技术来探索自组装途径（8-18个月）。3. 蛋白质笼中蛋白质自组装的研究（18-36个月）。1. 金属中心精确力场的发展。虽然存在大量用于有机分子的自动化力场（FF）拓扑构建器，但对过渡金属（tm）模型的系统扩展仍然具有挑战性。在这里，我们将利用机器学习（ML）来开发描述金属中心在水溶液中的行为的势。具体来说，我们将使用高斯近似势（GAP）框架，6结合从头算和密度泛函理论计算，来开发可以描述水相金属中心的ML力场势。为了测试这些模型的准确性，我们将评估实验数据可用的水络合物的性质。最初考虑的金属包括：Mg （II）、Fe （II）、Fe （III）、Cu（II）、Co（II）、Zn（II）和Ni（II）。2. 运用高效建模技术探索自组装途径。本课程将向学生介绍马尔可夫状态模型（MSM），以探索采用小肽系统的金属驱动自组装的动力学和动力学。核磁共振和量热实验数据将用于验证我们的结果。为此目的，将对软件包PyEMMA进行初步测试。7 3。蛋白质笼中蛋白质自组装的研究。在最后阶段，我们的目标是应用上述力场模型和采样技术来研究铁蛋白超家族成员的自组装。这些知识将有助于在新的工业相关应用中实现控制装配/拆卸。我们的目标是解决关于它们的组装途径，亚稳态物种的形成和稳定性，金属依赖性和再工程潜力的开放性问题。为了朝着这个目标迈出切实的一步，我们将首先研究低阶物种，这些物种在进入高阶聚集体或原子分辨率之前，预计会在组装过程中处于亚稳态。这将与David Clarke博士（爱丁堡）合作完成，他已经对该系统进行了表征，并为测试计算设计的结构提供了适当的协议。研究领域：理论化学

Metal ions play a vital role in many biological and chemical processes. Besides being of fundamental interest in chemistry, ions play a critical role in the assembly and function of biomolecules. Examples include protein cages,1, 2 multi-enzyme complexes,3 and RNA. Moreover, in combination with organic molecules, they also serve as building blocks for the assembly of synthetic systems metallocages. Metal-mediated self-assembly has emerged as a promising approach to develop complex biomimetic systems from simpler building blocks.4, 5 However, even for naturally occurring self-assemblies, understanding the fundamental rules driving self-organisation, structure and evolution remains a challenge, which substantially limits our ability to re-purpose the properties of these systems. The overall aim of this project is to . This aim will be met by addressing the following aspects: 1. Development of accurate force fields for metal centres (1-9 months). 2. Implementation of efficient modelling techniques to explore of self-assembly pathways (8-18 months). 3. Study of protein self-assembly in protein cages (18-36 months). 1. Development of accurate force fields for metal centres. While a plethora of automated force-field (FF) topology builders exist for organic molecules, systematic extensions to model transition metals (TMs) remain challenging. Here, we will utilise machine learning (ML) to develop potentials describing the behaviour of metal centres in aqueous solvent. Specifically, we will use the Gaussian Approximation Potentials (GAP) framework,6 in combination with ab initio and density functional theory calculations, to develop ML force field potentials that can describe metal centres in aqueous phase. To test the accuracy of these models, we will evaluate the properties of aquo-complexes for which experimental data are available. The metals initially considered will include: Mg (II), Fe (II), Fe (III), Cu(II), Co(II), Zn(II), and Ni(II). 2. Implementation of efficient modelling techniques to explore of self-assembly pathway. The student will be introduced to Markov state models (MSM) to explore the dynamics and kinetic of metal-driven self-assembly employing small peptide systems. NMR and calorimetry experimental data will be use to validate our results. The software package PyEMMA will be initially tested for this purpose.7 3. Study of protein self-assembly in protein cages. In a final stage, we aim to apply the force field models and the sampling techniques mentioned above to study the self-assembly in members of the ferritin superfamily. This knowledge will facilitate efforts towards controlled assembly/disassembly for new industrially relevant applications. We aim to address open questions regarding their assembly pathways, formation and stability of metastable species, metal-dependence and the potential for re-engineering. To make tangible steps towards this goal we will first study lower order species, which are predicted to be metastable states along the assembly process before moving into higher order aggregates or atomic resolution. This will be done in collaboration with Dr David Clarke (Edinburgh) who has carried out characterisation of this system and has the appropriate protocols for testing computationally designed constructs.Area: Theoretical Chemistry