SCOPE Dataset

该数据集是一个大规模、高质量的关注于人类药物-靶点相互作用（DTI）的数据集，它由13个公共数据库构建而成，大幅提升了数据量，并平衡了类别分布。该数据集包含了使用AlphaFold2对蛋白质和RDKit对化合物生成的结构信息，且相互作用被标记为阳性或阴性。与现有基准相比，该数据集的相互作用数据增加了20到100倍。其任务是对药物-靶点相互作用进行预测。

This is a large-scale, high-quality dataset focused on human drug-target interaction (DTI). It is constructed from 13 public databases, which substantially expands the dataset scale and balances the class distribution. This dataset contains structural information of proteins generated by AlphaFold2 and compounds generated by RDKit, and the interactions are labeled as positive or negative. Compared with existing benchmarks, the volume of interaction data in this dataset has increased by 20 to 100 times. The task of this dataset is to predict drug-target interactions.