Supporting Dataset for Simulations of High Temperature Decomposition of Metal-Organic Frameworks to form Amorphous Catalysts

This dataset contains computational data supporting the research article "Simulations of High Temperature Decomposition of Metal-Organic Frameworks to form Amorphous Catalysts." The repository includes training data, validation scripts, quench trajectories, and analysis scripts used in the study.File Format & Access The complete dataset is provided as a compressed .tar.xz archive. Users must extract this archive to access the directory structure and files described below.Content Overview Once extracted, the repository is organized into three primary directories covering foundational model validation, fine-tuning, and trajectory analysis:1-foundational_models/: Contains scripts for validating baseline foundational Machine Learning Interatomic Potentials (MLIPs), including MAE calculation utilities (e.g., validate.py, array_MAE.py).2-finetuning/: Includes the complete fine-tuning pipeline for MACE models. This directory houses:Training and validation datasets (.xyz format).Scripts for generating training sets and sampling structures (gen_training_sets.py, training_sampling.py).Fine-tuned and scratch MACE models.3-trajectories/: Contains data related to trajectory generation and analysis, including:Frameworks: Input structure files (CIF, XYZ) for frameworks (e.g., UiO-66, MIP-206) and guest molecules.Training & Finetuned Trajectories: Initial 1 ns quench simulations and trajectories generated using fine-tuned MLPs.Scripts: A comprehensive suite of Python scripts for running quench simulations, metadynamics, and post-processing analysis.For a detailed breakdown of the file tree and specific script descriptions, please refer to the README.md included within the root of the extracted archive.