Chen Changjun
Biomolecular Physics and Modeling Group, School of Physics, Huazhong University of Science and Technology , Wuhan 430074, Hubei, China.
J Phys Chem B. 2016 Mar 31;120(12):3061-71. doi: 10.1021/acs.jpcb.5b11892. Epub 2016 Mar 22.
The free energy landscape is the most important information in the study of the reaction mechanisms of the molecules. However, it is difficult to calculate. In a large collective variable space, a molecule must take a long time to obtain the sufficient sampling during the simulation. To save the calculation quantity, decreasing the sampling region and constructing the local free energy landscape is required in practice. However, the restricted region in the collective variable space may have an irregular shape. Simply restricting one or more collective variables of the molecule cannot satisfy the requirement. In this paper, we propose a modified tomographic method to perform the simulation. First, it divides the restricted region by some hyperplanes and connects the centers of hyperplanes together by a curve. Second, it forces the molecule to sample on the curve and the hyperplanes in the simulation and calculates the free energy data on them. Finally, all the free energy data are combined together to form the local free energy landscape. Without consideration of the area outside the restricted region, this free energy calculation can be more efficient. By this method, one can further optimize the path quickly in the collective variable space.
自由能景观是分子反应机制研究中最重要的信息。然而,它很难计算。在一个大的集体变量空间中,分子在模拟过程中必须花费很长时间才能获得足够的采样。为了节省计算量,在实际中需要减小采样区域并构建局部自由能景观。然而,集体变量空间中的受限区域可能具有不规则形状。简单地限制分子的一个或多个集体变量并不能满足要求。在本文中,我们提出了一种改进的断层扫描方法来进行模拟。首先,它用一些超平面划分受限区域,并通过一条曲线将超平面的中心连接起来。其次,它迫使分子在模拟中在曲线和超平面上进行采样,并计算它们上的自由能数据。最后,将所有自由能数据组合在一起形成局部自由能景观。不考虑受限区域之外的区域,这种自由能计算可以更高效。通过这种方法,人们可以在集体变量空间中更快地进一步优化路径。