This is for the oxadiazole NEB example. Basically the steps being automated are:
- Obtain PET-MAD
- Minimize endpoints and re-align endpoints (IRA+EON)
- Generate IDPP path (ASE)
- Run EON’s MMF-NEB (RO-NEB)
- Plot
Conceptually similar to the reproduction done for [1, 2] on Github (1, 2). We
use snakemake for this section[fn:: Some of these are inspired by the
workflows of amritagos].
Essentially this is as implemented with IRA for endpoint alignment [3] while the rest of the workflow uses from [4] in EON (also here), using systems from [5].
pixi s -e eon
snakemake -c4 # or c1
# View results/plots/04_plots/roneb.pngIf faster experimental visualization is required (compared to editing the config/general_config.yaml) then the results can be used directly:
cd results/03_neb
python -m rgpycrumbs.cli eon plt_neb --con-file neb.con --plot-structures "crit_points" --facecolor "floralwhite"[1] R. Goswami, M. Masterov, S. Kamath, A. Pena-Torres, and H. Jónsson, “Efficient Implementation of Gaussian Process Regression Accelerated Saddle Point Searches with Application to Molecular Reactions,” J. Chem. Theory Comput., Jul. 2025, doi: 10.1021/acs.jctc.5c00866.
[2] R. Goswami, “Bayesian hierarchical models for quantitative estimates for performance metrics applied to saddle search algorithms,” AIP Adv., vol. 15, no. 8, p. 85210, Aug. 2025, doi: 10.1063/5.0283639.
[3] M. Gunde, N. Salles, A. Hémeryck, and L. Martin-Samos, “IRA: a shape matching approach for recognition and comparison of generic atomic patterns,” J. Chem. Inf. Model., vol. 61, no. 11, pp. 5446–5457, Nov. 2021, doi: 10.1021/acs.jcim.1c00567.
[4] V. Ásgeirsson et al., “Nudged Elastic Band Method for Molecular Reactions Using Energy-Weighted Springs Combined with Eigenvector Following,” J. Chem. Theory Comput., vol. 17, no. 8, pp. 4929–4945, Aug. 2021, doi: 10.1021/acs.jctc.1c00462.
[5] O.-P. Koistinen, V. Ásgeirsson, A. Vehtari, and H. Jónsson, “Minimum Mode Saddle Point Searches Using Gaussian Process Regression with Inverse-Distance Covariance Function,” J. Chem. Theory Comput., vol. 16, no. 1, pp. 499–509, Jan. 2020, doi: 10.1021/acs.jctc.9b01038.
All the graphs can be re-generated via:
pixi r all-graphs
