膜蛋白处于复杂动态的生物膜环境中，导致单纯依靠结构信息难以完全诠释其行使生物学功能的微观机制。全原子分子动力学模拟已成为连接蛋白质结构与功能研究的有力工具，但由于膜蛋白功能涉及到复杂膜环境中蛋白质的长时间和大尺度动态行为，致使其计算效率非常低，难以广泛应用。尽管现有粗粒化分子模型的计算速度可达到上述研究要求，但由于粗粒化粒子的几何形状和力场参数过于简单和粗糙，使得其计算精度难以满足需求。基于我们前期工作基础，本项目拟针对标准氨基酸、不同种类磷脂、胆固醇和调控膜蛋白功能的重要有机分子，结合广义范德华势和多极矩展开静电势，建立具有全原子模拟精度的蛋白质和膜环境的新型粗粒化分子模型，并开展系统的理论研究。此模型的建立及相应计算程序的实现将帮助我们更加全面细致地理解膜蛋白在复杂生物膜环境中的结构、动力学行为及行使功能的微观机制，为复杂膜蛋白的功能研究以及相关药物研发提供强有力的理论支撑和计算工具。

The molecular mechanism of membrane protein functions cannot be explained easily with the structure information only, since membrane protein works under the complex and dynamic membrane environment. Fully atomistic molecular dynamics (MD) simulations provide a powerful tool to bridge the gap between structure and function studies. However, the functions of membrane proteins couple closely with their dynamic behaviors at long temporal and large spatial scale under complicated membrane environment, this results in much lower ower computational efficiency of fully atomistic MD simulations, and limits its widely uses for studying complicated membrane proteins. Although existed coarse grained models achieve the goal for above studies at the aspect of computing speed, they do not satisfy the accuracy requirement for membrane protein studies resulting from the oversimplification and rough description of geometric shape and force field parameters of their coarse grained particles. Based on our research experience and preliminary results, we propose to construct a new type of coarse grained model of proteins and its membrane environments with the same accuracy as fully atomistic modeling by combining generalized vdW potential and multipole expansion of electrostatic potential, which include standard amino acids, different types of lipids, cholesterol and important organic molecules which regulate the functions of membrane protein we will study. Meanwhile, we will use this model to do theoretical studies systematically. The construction of this new coarse grained model with fully atomistic modeling accuracy and implementation of its computer program will help us to understand deeply and precisely the properties of membrane proteins and the mechanism of their functions. This new computational models and techniques will build strong theoretical basis for membrane protein function studies, and offer a highly efficjent tool for membrane protein-targeted drug discovery.