面向计算机辅助分子机器设计的模拟方法和应用研究

Due to their small size, adaptive and controllable properties, molecular machines have promising potential applications in the fields of new materials, devices and molecular robots. The most important property of molecular machines are molecular motion. Movements observed in molecular machines are usually diverse and complex. Case in point, shuttling, rotation and isomerization, which are commonly viewed as independent motions, when, in reality, they are often coupled. Synthesis of such nanomolecular architectures by and large rests upon heuristics and the irreplaceable, longstanding experience of chemists. Under most circumstances, physical properties are determined post-facto, downstream from costly syntheses — sometimes revealing undesired features in the movements of the nano-objects. Such undesired features could be safely avoided by dint of a more efficient and rational strategy for the design of novel molecular machines endowed with specific functions, in a spirit similar to that of de novo drug design. Towards this end, theoretical prediction relying on robust concepts represents a most promising avenue. The research target of the present project is to puts forth a formal theoretical framework for revealing the mechanism of complex molecular motion in molecular machines and for evaluating the performance of motion. This theoretical framework will allow us to shed new light on the intricate nature of movements, and to explore the effects of structural modifications and to optimize external conditions to design molecular machines that have specific functions. The proposed methods are molecular dynamics simulations and free-energy calculations of the movement processes involved. According to the diversity and coupling of the molecular movements, we will put forward creative methods for high dimensional free-energy calculations. These new methods will be applied to the study of multi-way molecular switches, biomolecular springs, molecular engines, and rotaxane-based ionophores in different conditions, pH, light, and solvent. The research results are expected to lay a theoretical foundation for the ultimate goal of establishment of a computer-aided design system of intelligent molecular machines.

分子机器由于其体积小且具有自适应性和可操控性，在新型材料、器件和分子机器人等领域有着诱人的潜在应用。分子机器的重要特性是分子运动，观察到的运动有多种形式，如穿梭、旋转和构象变化，而这些看似独立的运动往往相互耦合。其物理性能的测定通常是在昂贵费时的合成之后，如何避免不希望的运动性能并实现对运动的调控是需要解决的关键问题。本项目的研究目标是借助计算机辅助药物设计的思想构建一个用于揭示分子机器中复杂分子运动的机理和评估运动性能的理论框架，并且能够通过优化结构和运动响应条件使得设计的分子机器具有特定的功能。拟采用的方法是分子动力学模拟和描述运动过程的自由能计算。针对分子运动的多样性和耦合性，将发展具有创新性的高效多维自由能计算新方法，并应用于不同控制条件（pH、光和溶剂）的多重分子开关，生物分子弹簧、分子引擎和基于轮烷的离子载体的研究，为最终建立计算机辅助智能分子机器设计系统奠定理论基础。

在扩展自适应偏置力（eABF）算法的基础上，实现了系列高效多维自由能计算新方法meta-eABF（eABF与metadynamics结合），WTM-eABF（引入well-tempering策略）和GaWTM-eABF（WTM-eABF与 GaMD结合），极大提高了采样效率、算法的渐进收敛性以及在正交空间中的采样效率。这些算法均已在国际流行的Colvars自由能计算模块中实现。进一步，在另一国际流行Plumed自由能计算模块实现了ELF子模块，使得eABF类方法通过ELF能够在多个分子动力学模拟软件中应用，使得eABF类方法能够在多个分子动力学模拟软件中应用。进一步实现了在复杂高维自由能面上寻找最优路径的高效算法MULE。将研发的高效自由能计算方法应用于多种不同控制条件的分子机器运动机制和功能的探索，揭示了基于分子梭构建的pH控制的荧光探针的多种功能和机理、多重分子开关中的单向运动的机理以及可能的调节策略；通过高维自由能计算对分子机器中相互耦合的多种运动进行了深入研究，如旋转、翻转、穿梭和构象变化等，揭示了分子机器中各种运动的协同对其性能影响的重要作用，以及水在分子机器运动中的润滑和阻尼效应。结果表明本项研究所提出的高效高维自由能计算方法和一套理论框架可用于计算机辅助分子器件设计。