New Generation DelPhi: large systems and beyond electrostatics

 描述（申请人提供）：分子生物学的主要对象是蛋白质、DNA和RNA，以及沿着各种小分子和大分子集合体。这些物体经常与纳米粒子一起参与纳米科学中的各种现象。随着实验和计算方法的发展，现在的研究人员正在通过研究微管，病毒和细胞器等系统的生物学特性来扩展其功能。这样的系统提出了两个主要的挑战：（a）经常他们的原子结构是无法通过实验获得的，必须进行建模;（B）他们有超过1,000 μ m的大尺寸，这是不能处理的大多数现有的建模软件包。通过该提案，我们计划解决这些挑战：（a）进一步扩展蛋白质纳米对象集成器（ProNOI）的功能，其允许从实验图像（如Cryo-EM图像）跟踪的对象的原子风格建模;（B）在RAM使用和计算速度方面扩展德尔菲功能，以允许对大尺寸系统进行常规建模。此外，上一个资助期的工作产生了面向对象的C++德尔菲代码，其中一个核心组件是有限差分（FD）算法。由于FD是求解微分方程（DE）的最通用的数值技术之一，我们将开发插件来解决DE描述不同于静电的量（如温度，热量，离子密度），并将使社区驱动的研究包括生物物理社区感兴趣的其他量的建模。此外，研究人员经常希望模拟由大分子与大细胞组分（如微管）相互作用组成的系统，同时探索不同的蛋白质取向和蛋白质与其结合的结合模式。为了促进这样的研究，我们将扩展现有的德尔菲聚焦技术的能力，以允许具有离网格中心和非互易尺度的父子聚焦运行。我们将使用一个模块来补充代码，该模块能够自动生成感兴趣的结构域/蛋白质的替代位置和方向，计算它们的能量并利用蒙特卡罗模拟程序来评估它们的概率。

 DESCRIPTION (provided by applicant): The main objects in molecular biology are proteins, DNAs and RNAs, along with various small molecules and large macromolecular assemblages. These objects are frequently involved in various phenomena in nano-science together with nano-particles. With the progress of both experimental and computational approaches, nowadays researchers are expanding the repertoire by investigating biological characteristics of systems like microtubules, viruses and cellular organelles. Such systems are posing two major challenges: (a) frequently their atomic structures are not experimentally available and have to be modeled; and (b) they have large dimensions above 1,000 Å, which cannot be handled by most of the existing modeling packages. With this proposal we plan to address these challenges: (a) further expand the capabilities of Protein-Nano Object Integrator (ProNOI) which allows for atomic style modeling of objects traced from experimental images (as Cryo-EM image); (b) expand DelPhi capabilities, in terms of RAM usage and speed of calculations, to allow systems with large dimensions to be modeled routinely. Furthermore, the work from the previous funding period resulted in object-oriented C++ DelPhi code and one of the core component is the finite-difference (FD) algorithm. Since the FD is one of the most universal numerical technique of solving differential equations (DE), we will develop plugins to solve DE describing quantities different from electrostatics (such as temperature, heat, ion density) and will enable community-driven research to include modeling of other quantities of interest for the biophysical community. Furthermore, frequently researchers want to model systems comprised of a macromolecule interacting with large cellular component (such as microtubule), while exploring different protein orientations and binding modes of the protein binding to it. To facilitate such a research, we will expand the capability of existing DelPhi focusing technique to allow for parent-son focusing runs with off-grid centering and nonreciprocal scale. We will complement the code with a module capable of automatically generating alternative positions and orientations of the domain/protein of interest, computing their energies and utilizing Monte Carlo simulation procedure to assess their probabilities.