Incorporating Linear Synchronous Transit Interpolation into the Growing String Method: Algorithm and Applications
Incorporating Linear Synchronous Transit Interpolation into the Growing String Method: Algorithm and Applications
复制标题
DOI：
10.1021/ct200654u
复制
发表时间：
2011-12-01
影响因子：
5.5
通讯作者：
Head-Gordon, Martin
中科院分区：
文献类型：
--
作者：
Behn, Andrew;Zimmerman, Paul M.;Head-Gordon, Martin
增长弦法是用理论方法系统研究化学反应的有力工具，可以快速识别连接已知反应物和产物结构的过渡态。然而，该方法的效率在很大程度上受到优化过程中向字符串添加新节点时插值方案的选择的影响。特别是，使用笛卡尔坐标与三次样条插值通常会产生远离最终反应路径的猜测结构，并且需要许多优化步骤（因此需要许多能量和梯度计算）才能产生合理的最终结构。在本文中，我们提出了一种使用 Halgren 和 Lipscomb 的线性同步传输方法在增长弦方法中对节点进行插值和重新参数化的新方法。当应用于丙氨酸二肽重排和简化的阳离子烷基环缩合反应时，计算成本显着加快（30-50%）。
The growing string method is a powerful tool in the systematic study of chemical reactions with theoretical methods which allows for the rapid identification of transition states connecting known reactant and product structures. However, the efficiency of this method is heavily influenced by the choice of interpolation scheme when adding new nodes to the string during optimization. In particular, the use of Cartesian coordinates with cubic spline interpolation often produces guess structures which are far from the final reaction path and require many optimization steps (and thus many energy and gradient calculations) to yield a reasonable final structure. In this paper, we present a new method for interpolating and reparameterizing nodes within the growing string method using the linear synchronous transit method of Halgren and Lipscomb. When applied to the alanine dipeptide rearrangement and a simplified cationic alkyl ring condensation reaction, a significant speedup in terms of computational cost is achieved (30-50%).