MS-ML/synth1_2x4.7M
收藏Hugging Face2025-09-19 更新2025-02-15 收录
下载链接：
https://hf-mirror.com/datasets/MS-ML/synth1_2x4.7M
下载链接
链接失效反馈官方服务：
资源简介：
这是一个由NEIMS和RASSP模型生成的GC-EI-MS合成数据集，包含4.7百万个独特的化合物和9.4百万个独特的光谱。数据集由从ZINC数据库中下载的4.7百万个结构生成，分为训练集、验证集和测试集，每个集合包含相同化合物但光谱来源不同。数据以JSONL格式存储，每行包含一个JSON对象，包括光谱强度、质量电荷比和SMILES字符串。
This is a GC-EI-MS synthetic dataset generated by NEIMS and RASSP models, containing 4.7 million unique compounds and 9.4 million unique spectra. The dataset is derived from 4.7 million structures downloaded from the ZINC database, split into training, validation, and test sets with the same compounds but different sources of spectra. The data is stored in JSONL format, with each line containing a JSON object that includes spectral intensity, mass-to-charge ratio, and SMILES strings.
提供机构：
MS-ML
