Transition1x

该数据集名为Transition1x，包含了960万个分子构型在反应路径上及周围区域的密度泛函理论（DFT）计算得到的力和能量数据。数据集的结构分为训练、测试和验证数据组，每组包含了每个反应的原子序数、能量、力和位置信息。该数据集的任务是用于训练在分子动力学中搜索反应能垒的机器学习模型。

This dataset, named Transition1x, contains force and energy data computed via density functional theory (DFT) for 9.6 million molecular configurations along and in the vicinity of reaction pathways. The dataset is divided into training, test, and validation subsets, each containing atomic numbers, energies, forces, and positional information for individual reactions. The core application of this dataset is to train machine learning models for identifying reaction energy barriers in molecular dynamics simulations.