利用 Kirkwood-Buff 分析方法评价和改进 CHARMM 力场参数

Empirical force field is the foundation for condensed phase simulation and computational biology. CHARMM force field, pioneered by Prof. Martin Karplus, the laureate for 2013 Nobel Chemistry Prize, is one of the most widely used molecular force fields. From late 1970s to the present, CHARMM as a set of force field has been almost completed, covering from small organic molecules, short peptides, to macromolecules such as proteins and nucleic acids. During the development of CHARMM, the developers employed quantum chemistry calculations, structural data from NMR and crystallography and other condensed phase experimental data for pure liquids in order to calibrate the parameters so that results from the simulations using those parameters are able to reproduce experimental data as accurate as possible. However, experimental data from liquid mixtures have never been part of CHARMM force field development. Lately, it was found that without including such experimental data explicitly into force field development, discrepancies exist under certain circumstances and sometimes such discrepancies can be quite large. Therefore, taking experimental data for liquid mixtures into account has become imperative. Kirkwood-Buff analysis is a powerful method to identify the differences between simulations and experiments. It is the goal of this proposal to use experimental data for liquid mixtures in CHARMM force field development and make it more accurate in reproducing experimental results from molecular dynamics simulations.

经验分子力场是凝聚态模拟和计算生物学的基础。2013年诺贝尔化学奖得主哈佛大学化学系教授 Martin Karplus 领导开发的CHARMM分子力场是现在广泛使用的分子力场之一。从1970年代末到现在，CHARMM 的发展日趋完善，力场参数涵盖了从小分子化合物、多肽到蛋白质和核苷酸。长久以来，在CHARMM 的开发过程中，开发者们广泛使用了量子化学的计算、各种晶体结构数据和其他的实验参数，使得力场参数尽可能的准确，在模拟过程中能够复制实验数据。但是，混合溶液的实验数据从来没有被用在CHARMM 力场开发的过程中。最近发现，忽略这样的实验数据，在模拟混合溶液时，会导致较大的偏差。应用 Kirkwoob-Buff （KB）的分析方法，能够找出这样的偏差所在。将混合溶液的实验数据加入到力场开发之中，辅以KB 分析方法，更新 CHARMM 力场参数将会更加准确的使得分子动力学模拟重复出实验数据。