Learning reaction coordinates via cross-entropy minimization: Application to alanine dipeptide
Learning reaction coordinates via cross-entropy minimization: Application to alanine dipeptide
复制标题
DOI：
10.1063/5.0009066
复制
发表时间：
2020-08-07
影响因子：
4.4
通讯作者：
Matubayasi, Nobuyuki
中科院分区：
文献类型：
--
作者：
Mori, Yusuke;Okazaki, Kei-ichi;Matubayasi, Nobuyuki
提出了一种从复杂分子系统中大量集体变量中求反应坐标的交叉熵最小化方法。这种方法是用s形描述提交者函数的似然最大化方法的扩展。通过设计，反应坐标作为各种集体变量的函数进行了优化，使得分子动力学模拟生成的提交者pB*值的分布可以以s型方式描述。我们还引入了用于机器学习领域的l -2范数正则化，以防止在考虑的集体变量数量很大时过拟合。本方法以45个二面角为候选变量，对丙氨酸二肽的真空异构化进行了研究。正则化参数通过使用训练和测试数据集的交叉验证来确定。结果表明，最优反应坐标包含重要的二面角，与前人的研究结果一致。此外，pB*近似于0.5的点清楚地表明，利用提取的二面角，在平均力势上有一个区分反应物和生成物状态的分离矩阵。
We propose a cross-entropy minimization method for finding the reaction coordinate from a large number of collective variables in complex molecular systems. This method is an extension of the likelihood maximization approach describing the committor function with a sigmoid. By design, the reaction coordinate as a function of various collective variables is optimized such that the distribution of the committor pB* values generated from molecular dynamics simulations can be described in a sigmoidal manner. We also introduce the L-2-norm regularization used in the machine learning field to prevent overfitting when the number of considered collective variables is large. The current method is applied to study the isomerization of alanine dipeptide in vacuum, where 45 dihedral angles are used as candidate variables. The regularization parameter is determined by cross-validation using training and test datasets. It is demonstrated that the optimal reaction coordinate involves important dihedral angles, which are consistent with the previously reported results. Furthermore, the points with pB*similar to 0.5 clearly indicate a separatrix distinguishing reactant and product states on the potential of mean force using the extracted dihedral angles.