Reproducibility of "Diffusion-based Generative AI for Exploring Transition States from 2D Molecular Graphs"

This file is the source data to ensure reproducibility of the paper "Diffusion-based Generative AI for Exploring Transition States from 2D Molecular Graphs". It contains the logs and results of all DFT calculations associated with transition states generated using the model proposed in the paper. It also includes code to reproduce the core findings of the paper, which can be done by running reproduce.sh. To accurately reproduce the results of the paper, use the v1.0.0 virtual environment from "https://github.com/seonghann/tsdiff".

本文件为保障论文《基于扩散模型的生成式AI：从二维分子图探索过渡态》（Diffusion-based Generative AI for Exploring Transition States from 2D Molecular Graphs）的可复现性提供源数据集。其收录了利用本文提出的模型生成的全部过渡态所对应的密度泛函理论（Density Functional Theory, DFT）计算日志与结果。本数据集同时附带复现该论文核心研究结论的代码，执行reproduce.sh脚本即可完成复现。若需精准复现该论文的实验结果，请使用来自"https://github.com/seonghann/tsdiff"的v1.0.0虚拟环境。