Durmaz Vedat
Zuse-Institut Berlin, Takustraße 7, Berlin, 14195, Germany,
J Comput Aided Mol Des. 2015 Mar;29(3):225-32. doi: 10.1007/s10822-014-9817-0. Epub 2014 Nov 27.
A conventional by hand construction and parameterization of a polymer model for the purpose of molecular simulations can quickly become very work-intensive and time-consuming. Using the example of polyglycerol, I present a polymer decomposition strategy yielding a set of five monomeric residues that are convenient for an instantaneous assembly and subsequent force field simulation of a polyglycerol polymer model. Force field parameters have been developed in accordance with the classical Amber force field. Partial charges of each unit were fitted to the electrostatic potential using quantum-chemical methods and slightly modified in order to guarantee a neutral total polymer charge. In contrast to similarly constructed models of amino acid and nucleotide sequences, the glycerol building blocks may yield an arbitrary degree of bifurcations depending on the underlying probabilistic model. The iterative development of the overall structure as well as the relation of linear to branching units is controlled by a simple Markov model which is presented with few algorithmic details. The resulting polymer is highly suitable for classical explicit water molecular dynamics simulations on the atomistic level after a structural relaxation step. Moreover, the decomposition strategy presented here can easily be adopted to many other (co)polymers.
为了进行分子模拟,采用传统的手工构建和聚合物模型参数化方法会很快变得工作量极大且耗时。以聚甘油为例,我提出了一种聚合物分解策略,该策略可产生一组五个单体残基,便于聚甘油聚合物模型的即时组装和后续力场模拟。力场参数是根据经典的Amber力场开发的。使用量子化学方法将每个单元的部分电荷拟合到静电势,并进行了轻微修改,以确保聚合物总电荷为中性。与类似构建的氨基酸和核苷酸序列模型不同,甘油构建块可能会根据基础概率模型产生任意程度的分支。整体结构的迭代开发以及线性单元与分支单元的关系由一个简单的马尔可夫模型控制,文中给出了该模型的一些算法细节。经过结构弛豫步骤后,所得聚合物非常适合在原子水平上进行经典的显式水分子动力学模拟。此外,这里提出的分解策略可以很容易地应用于许多其他(共)聚合物。