Achieving Long Timescale Sampling in Biomolecular Simulations

该奖项是根据2009年美国复苏和再投资法案（公法111-5）资助的。本研究的目的是开发和应用有效的模拟方法来定量描述复杂的生物分子现象，这些现象通常发生在比传统方法更长的时间尺度上。分子识别，如蛋白质-配体识别和大构象转变是生物分子机器的两个关键要素。由于时间尺度的限制，通过计算机模拟来定量理解这些事件非常具有挑战性。在本项目中，将开发和应用高效的广义集成模拟方法来提高生物分子模拟的采样能力。这个项目的前景特别得到了PI实验室最近方法论发展的保证：正交空间随机漫步（OSRW）算法，它可以以鲁棒和有效的方式独特地实现集体环境放松。具体而言，将有三个目标：(1)通过基于OSRW的自由能模拟来理解蛋白质-蛋白质相互作用和蛋白质静电；(2)开发基于混合量子力学/分子力学（QM/MM）的OSRW采样方案，以更准确地描述蛋白质与配体的相互作用；(3)发展基于OSRW的构象采样方法预测蛋白质功能动力学。近年来，PI在组织研讨会和会议方面做出了巨大的努力，以连接计算生物物理学和实验生物物理学，并在方法发展和生物学应用之间架起桥梁。其中一些讲习班，例如关于加强抽样的碲化物讲习班，已成为常设讲习班。特别是，PI一直非常重视为年轻科学家建立多层次的环境，使他们能够接受适当的教育，进行多学科研究。为了使PI和其他团队最近开发的先进方法迅速被科学界吸收，PI于2009年初在《计算化学杂志》上编辑了一期关于自由能模拟的特刊，并与人合作撰写了一本专注于现代模拟方法及其实际应用的书。此外，通过该项目开发的方法将在CHARMM程序中发布，可广泛应用于一般科学界。

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).The objective of this research is to develop and apply efficient simulation methods to quantitatively describe complex biomolecular phenomena that usually occur in longer timescale than what conventional methods can achieve. Molecular recognition such as protein-ligand recognition and large conformational transition are two key elements in biomolecular machines. Due to timescale limitation, quantitative understanding of these events via computer simulation has been very challenging. In this project, efficient generalized ensemble simulation methods will be developed and employed to enhance the sampling power of biomolecular simulations. The promise of this project is particularly ensured by a recent methodological development in the PI's lab: the orthogonal space random walk (OSRW) algorithm, which can uniquely realize collective environmental relaxation in a robust and efficient manner. Specifically, three aims will be targeted: (1) understanding protein-protein interactions and protein electrostatics via OSRW based free energy simulations; (2) developing a hybrid quantum mechanical/molecular mechanical (QM/MM) based OSRW sampling scheme to more accurately describe protein-ligand interactions; (3) developing OSRW based conformational sampling methods to predict protein functional dynamics. In recent years, the PI has made tremendous efforts in organizing workshops and meetings to connect computational biophysics and experimental biophysics and to bridge method developments and biological applications. Some of these workshops, for instance, Telluride workshop on sampling enhancement, have become standing ones. In particular, the PI has been putting enormous emphasis in building multi-layered environments for young scientists so that they can receive suitable education to carry out multi-disciplinary research. In order for the advanced methods that are recently developed in the groups of PI and others to be quickly absorbed by the scientific community, the PI edited a special issue on free energy simulation in Journal of Computational Chemistry in early 2009 and is co-writing a book that is focused on modern simulation methods and their hand-on application. In addition, the developed methods through this project will be released in the program CHARMM, which can be widely employed by general scientific community.