新型多重迭代自适应播种模拟方法及其在蛋白质折叠与构象转换上的应用

Molecular Dynamics (MD) simulation has become an important method to study structures and dynamics of proteins, however, a large challenge has been faced: the sampling efficiency of Conventional MD is low while Enhanced MD can’t provide real dynamic characteristics. To overcome the challenge, in this project, based on the adaptive seeding method, a new idea of feedback regulation on conformational transition networks will be proposed, and a new simulation method with multiple iterations of adaptive seeding, which combines Enhanced MD, Conventional MD as well as the Markov state model, will be developed to raise efficiency and provide real dynamic characteristics simultaneously; moreover, this new method will be applied to study two core problems of proteins: folding and conformational transition of proteins. This project will provide a deep investigation into transition networks and free energy barriers during the processes of folding and conformational transition of proteins, and thus promote our understanding of formation of protein structures, protein function and molecular mechanism of relevant biological processes, its achievements will have potential but profound theoretical significance for revealing relevant pathogenesis such as parkinsonism, BSE, etc, as well as developing relevant drugs.

分子动力学模拟已成为研究蛋白质结构及动力学的重要手段，然而也面临较大挑战：常规分子动力学模拟的采样效率较低，而增强采样分子动力学模拟无法获得真实的动力学特性。为克服这一挑战，本项目计划以自适应播种算法为基础，针对构象转换网络提出反馈调节思想，开发一套增强采样分子动力学模拟、常规分子动力学模拟与马尔科夫状态模型分析高度结合的多重迭代自适应模拟方法，在保持真实动力学特性的同时，提高采样效率；并将该方法运用于当今蛋白质领域的两大核心问题：蛋白质折叠以及蛋白质构象转换上。本项目的实施，将深入研究蛋白质折叠过程以及蛋白质构象转换过程的转换网络以及自由能垒等，从而帮助人们全面理解蛋白质结构的形成、功能以及相关生物学过程的分子机制；相关的研究成果对揭示帕金森症、疯牛病等相关重大疾病的致病机理，以及相关药物的开发具有潜在而深远的理论指导意义。

当今理论生物物理领域所面临的一个重大难题是如何在提高分子动力学模拟采样效率的同时，保留真实的动力学特性。分子动力学模拟已成为研究生物大分子构象转换的重要手段，然而在理论上也面临重大挑战：常规分子动力学模拟的采样效率较低，而增强采样分子动力学模拟无法获得真实的动力学特性，效率与精度难以兼顾。为克服这一挑战，本项目将以马尔科夫状态模型以及自适应播种算法为基础，提出针对构象转换网络的反馈调节思想，开发一套增强采样分子动力学模拟、常规分子动力学模拟与马尔科夫状态模型分析高度结合的新型多重自适应播种模拟方法，达到效率与精度并举，并应用于具有重要生物学意义的HPPK 蛋白酶体系。本项目的实施，将为深入研究生物大分子不同构象状态之间的转换网络、自由能垒等提供重要的理论模拟方法，从而帮助人们全面理解生物大分子功能的微观机理；相关研究成果对相关致病机理的揭示，以及相关药物的开发具有潜在而深远的理论指导意义。