Characterizing Reaction Route Map of Realistic Molecular Reactions based on Weight Rank Clique Filtration of Persistent Homology
arXiv:2211.15067 · doi:10.1021/acs.jctc.2c01204
Abstract
A reaction route map (RRM) constructed using the GRRM program is a collection of elementary reaction pathways, each of which comprises two equilibrium (EQ) geometries and one transition state (TS) geometry connected by an intrinsic reaction coordinate (IRC). An RRM can be mathematically represented by a graph with weights assigned to both vertices, corresponding to EQs, and edges, corresponding to TSs, representing the corresponding energies. In this study, we propose a method to extract topological descriptors of a weighted graph representing an RRM based on persistent homology (PH). The work of Mirth et al. [J. Chem. Phys. 2021, 154, 114114], in which PH analysis was applied to the (3N-6)-dimensional potential energy surface of an N atomic system, is related to the present method, but our method is practically applicable to realistic molecular reactions. Numerical assessments revealed that our method can extract the same information as the method proposed by Mirth et al. for the 0-th and 1-st PHs, except for the death of the 1-st PH. In addition, the information obtained from the 0-th PH corresponds to the analysis using the disconnectivity graph. The results of this study suggest that the descriptors obtained using the proposed method accurately reflect the characteristics of the chemical reactions and/or physicochemical properties of the system.
arXiv comment46 pages, 21 figures
References in corpus (7)
- A roadmap for the computation of persistent homology
- Topological strata of weighted complex networks
- The Exploration of Chemical Reaction Networks
- Chemoton 2.0: Autonomous Exploration of Chemical Reaction Networks
- Molassembler: Molecular graph construction, modification and conformer generation for inorganic and organic molecules
- Pathfinder -- Navigating and Analyzing Chemical Reaction Networks with an Efficient Graph-based Approach
- Representations of Energy Landscapes by Sublevelset Persistent Homology: An Example With n-Alkanes