Supporting Code for “Equivariant Neural Networks Reveal How Host–Guest Interactions Shape ¹²⁹Xe NMR in Porous Liquids”

Graphical Abstract

---

📄 Author: Ouail Zakary

• 📧 Email: Ouail.Zakary@oulu.fi
• 🔗 ORCID: 0000-0002-7793-3306
• 🌐 Website: Personal Webpage
• 📁 Portfolio: Academic Portfolio

---

This is the supporting code for the manuscript “Equivariant Neural Networks Reveal How Host–Guest Interactions Shape ¹²⁹Xe NMR in Porous Liquids”.

The repository contains the following sections:

1. Code for dataset preparation for MLIP:
   i. Configuration generation using semi-empirical MD (SEMD) simulations. (directory)
   ii. Single-point DFT calculations of the sampled SEMD snapshots. (directory)
   iii. Dataset formatting for Allegro architecture. (directory)
2. Code for principal component analysis (PCA). (directory)
3. Code for t-distributed stochastic neighbor embedding (t-SNE). (directory)
4. Code for training, validation, and testing of Allegro architecture. (directory)
5. Code for machine learning MD simulations. (directory)
6. Code for dataset preparation for the NMR-ML model:
   i. Configuration generation using SEMD simulations. (directory)
   ii. DFT calculations of ¹²⁹Xe NMR magnetic shielding tensor, σ. (directory)
   iii. Dataset formatting for MatTen architecture. (directory)
7. Code for training, validation, and testing of MatTen. (directory)
8. Prediction of ¹²⁹Xe σ from the pre-trained NMR-ML model. (directory)
9. Python scripts and raw numerical data for all figures in the main manuscript and the supporting information. (directory)

Citations

If you use the code in this repository, please cite the following:

Paper

@article{zakary_xe-at-cc3-tba_2025,
  title={Equivariant Neural Networks Reveal How Host--Guest Interactions Shape $^{129}$Xe NMR in Porous Liquids},
  author={Zakary, Ouail and Lantto, Perttu},
  journal={The Journal of Physical Chemistry Letters},
  volume={16},
  pages={12095--12103}, 
  year={2025},
  publisher={American Chemical Society},
  doi={10.1021/acs.jpclett.5c02846},
  url={https://doi.org/10.1021/acs.jpclett.5c02846}
}

Dataset

@dataset{zakary_xe-at-cc3-tba_2025,
  author = {Zakary, Ouail and Lantto, Perttu},
  title = {Supporting Data for "Equivariant Neural Networks Reveal How Host--Guest Interactions Shape $^{129}$Xe NMR in Porous Liquids"},
  year = {2025},
  publisher = {Zenodo},
  doi = {10.5281/zenodo.17105321},
  url = {https://doi.org/10.5281/zenodo.17105321}
}

Code

@misc{zakary_xe-at-cc3-tba_github_2025,
  author = {Zakary, Ouail and Lantto, Perttu},
  title = {Supporting Code for "Equivariant Neural Networks Reveal How Host--Guest Interactions Shape $^{129}$Xe NMR in Porous Liquids"},
  year = {2025},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/ozakary/data-Xe_at_CC3_at_TBA}},
  url = {https://github.com/ozakary/data-Xe_at_CC3_at_TBA}
}

---

For further details, please refer to the respective folders or contact the author via the provided email.