Alchemy
收藏资源简介:
Alchemy数据集是由腾讯等机构创建,专注于化学和材料科学的机器学习模型开发。该数据集包含119,487个有机分子,每个分子具有多达14个重原子,涵盖12种量子力学属性。数据来源于GDB MedChem数据库,通过PySCF计算框架计算得到。Alchemy数据集不仅扩展了现有分子数据集的体积和多样性,还特别关注药物化学领域,旨在通过提供高质量数据促进机器学习模型在化学和材料科学中的应用,解决分子属性预测等关键问题。
The Alchemy dataset was created by Tencent and other institutions, focusing on the development of machine learning models for chemistry and materials science. It contains 119,487 organic molecules, each with up to 14 heavy atoms, covering 12 quantum mechanical properties. The data is derived from the GDB MedChem database and calculated via the PySCF computational framework. The Alchemy dataset not only expands the scale and diversity of existing molecular datasets but also pays special attention to the field of medicinal chemistry, aiming to promote the application of machine learning models in chemistry and materials science by providing high-quality data and addressing key issues such as molecular property prediction.

- 1Alchemy: A Quantum Chemistry Dataset for Benchmarking AI Models腾讯1,香港中文大学2,多伦多大学3,清华大学4,向量研究所5 · 2019年



