scikit-fingerprints/MoleculeNet_MUV
收藏Hugging Face2024-07-18 更新2024-07-22 收录
下载链接:
https://hf-mirror.com/datasets/scikit-fingerprints/MoleculeNet_MUV
下载链接
链接失效反馈官方服务:
资源简介:
MoleculeNet MUV数据集是MoleculeNet基准测试的一部分,用于虚拟筛选技术的验证。该数据集包含17个二元分类任务,基于PubChem BioAssays设计。总样本量为93087,推荐使用scaffold分割方法,并建议使用AUPRC和AUROC作为评估指标。需要注意的是,目标值存在缺失值,算法应仅在现有标签上进行评估,训练数据可能需要用零值进行填补。
The MoleculeNet MUV dataset is part of the MoleculeNet benchmark, designed for validating virtual screening techniques. It contains 93087 samples for predicting 17 targets based on PubChem bioassays. All tasks are binary classification tasks, and targets include missing values. AUPRC and AUROC are recommended as evaluation metrics.
提供机构:
scikit-fingerprints



