HydroNet
收藏资源简介:
HydroNet数据集是由太平洋西北国家实验室创建的,包含495万个水分子团簇的数据集。该数据集提供了空间坐标和两种类型的图表示,以适应各种机器学习实践。数据集通过Monte Carlo Temperature Basin Paving (MCTBP) 采样方法生成,覆盖了3至30个水分子每团簇的低能量水分子团簇。HydroNet数据集主要用于机器学习在化学领域的应用,特别是预测化学性质和生成具有定制性质的分子结构,旨在解决分子间和/或长程相互作用的问题。
The HydroNet dataset, developed by the Pacific Northwest National Laboratory, contains 4.95 million water molecule clusters. This dataset provides spatial coordinates and two types of graph representations to support diverse machine learning workflows. Generated via the Monte Carlo Temperature Basin Paving (MCTBP) sampling method, it covers low-energy water clusters consisting of 3 to 30 water molecules per cluster. Primarily applied to machine learning research in chemistry, the HydroNet dataset is specifically utilized for predicting chemical properties and generating molecular structures with tailored properties, with the goal of addressing challenges associated with intermolecular and/or long-range interactions.

- 1HydroNet: Benchmark Tasks for Preserving Intermolecular Interactions and Structural Motifs in Predictive and Generative Models for Molecular Data太平洋西北国家实验室 · 2020年



