Efficient Simulation of Protein-Membrane Interactions by Implicit Solvent Algorithms

本项目旨在开发一种新的计算模型来研究膜相互作用下蛋白质构象的变化。其目标是：(1)开发一种鲁棒且高效的并行迭代算法来求解泊松-玻尔兹曼方程；(2)通过大量减少特定蛋白质家族的构象自由变量的数量来定义新的生物分子势能函数；(3)开发了一种新的稀疏矩阵压缩方案，用于以最优内存位置顺序编程高效的应用定制预调节器；(4)建立约束分子动力学模拟的并行迭代算法；(5)培养面向后基因组时代的本科生和研究生研究人员。具体而言，新算法和新模型将应用于研究蛋白毒素Cyt1A在膜相互作用下引起膜损伤的分子机制。理论结果还将与生化和生物物理方法（如荧光光谱、表面张力测定法和电子显微镜）产生的实验室数据进行比较。计算和实验数据的综合可能为不穿过脂质双分子层的蛋白质对膜的渗透和损伤方式带来新的见解。在生物技术、医学、农业、食品、制药或国防工业中，有效的算法和蛋白质-膜相互作用的建模可以节省大量的时间、金钱和实验室工作。本项目开发的新算法和新模型可以显著简化大型蛋白质-水膜系统模拟的计算复杂度。它们也可以应用于各种生物分子模拟。事实上，泊松-玻尔兹曼方程的数值解在隐式溶剂模型中起着核心作用。分子势能函数是分子建模的基础，分子势能函数的最小化和分子动力学模拟是计算生物学的两项基本任务。此外，所选择的模型蛋白，毒素Cyt1A，已被用作一种环境安全的杀虫剂，专门针对蚊子和黑蝇。阐明其作用方式可能有助于业界通过靶向特异性和与其他杀虫剂的协同作用来提高毒素的功效。此外，通过数学家/计算机科学家和生物化学家/生物物理学家之间的合作，参与该项目的学生将在数学、计算机科学和生命科学的虚拟中心接受培训。这项资助是在DMS/NIGMS联合倡议下提供的，以支持数学生物学领域的研究资助。这是一个由美国国家科学基金会数学科学部（DMS）和美国国立卫生研究院国家综合医学科学研究所（NIGMS）联合主办的竞赛。

This project aims to develop a new computational model for studying protein conformation changes upon membrane interactions. Its objectives are to (1) develop a robust and efficient parallel iterative algorithm for solving the Poisson-Boltzmann equation; (2) define a new biomolecular potential-energy function through a large reduction of the number of conformation-freedom variables for a family of particular proteins; (3) develop a new sparse matrix compress scheme for programming an efficient application-tailored preconditioner in an optimal order of memory locations; (4) develop a parallel iterative algorithm for constrained molecular dynamics simulations; and (5) train undergraduate and graduate researchers for the post-genomic era. Specifically, the new algorithms and the new model will be applied to studying the molecular mechanism of membrane damage induced by the protein toxin Cyt1A upon membrane interactions. The theoretical results will also be compared to laboratory data produced by biochemical and biophysical methods such as fluorescence spectroscopy, surface tensiometry, and electron microscopy. Synthesis of the computational and experimental data may bring new insights to the manner of membrane permeabilization and damage by proteins that do not transverse the lipid bilayer. Efficient algorithms and modeling of protein-membrane interactions can save considerable amounts of time, money, and laboratory effort in biotechnology, medicine, agriculture, and the food, pharmaceutical, or defense industries. The new algorithms and the new model developed in this project can significantly simplify computing complexity in simulations of large protein-water-membrane systems. They can also be applied to various biomolecular simulations. In fact, numerical solution of the Poisson-Boltzmann equation plays a central role in implicit solvent models. The molecular potential energy function is a cornerstone of molecular modeling, and the minimization of the molecular potential energy function and the molecular dynamics simulation are two fundamental tasks in computational biology. Furthermore, the chosen model protein, the toxin Cyt1A, has been used as an environmentally-safe insecticide specific against mosquitoes and blackflies. Elucidating its mode of action may help the industry increase the toxin's efficacy by targeting specificity and synergism with other insecticides. Additionally, through the collaboration between a mathematician/computer scientist and a biochemist/biophysicist, students involved in this project will be trained in a virtual hub of mathematics, computer science, and the life sciences. This grant is made under the Joint DMS/NIGMS Initiative to Support Research Grants in the Area of Mathematical Biology. This is a joint competition sponsored by the Division of Mathematical Sciences (DMS) at the National Science Foundation and the National Institute of General Medical Sciences (NIGMS) at the National Institutes of Health.