Bindwell/binding-affinity-PL

Binding Affinity数据集是一个综合性的蛋白质-配体结合亲和力数据集合，包含了多个来源的数据。数据集结构包含多个分割，每个分割对应一个特定的数据源，如BindingDB IC50、BindingDB Kd、Davis、KIBA、PDBBind 2020等。此外，还提供了一个主训练分割，该分割聚合了所有来源的数据。数据集经过了合并、去重和归一化等预处理步骤，以确保数据的唯一性和一致性。

The Binding Affinity dataset is a comprehensive collection of protein-ligand binding affinity data, compiled from multiple sources. The dataset is structured with multiple splits, each corresponding to a specific source such as BindingDB IC50, BindingDB Kd, Davis, KIBA, PDBBind 2020, etc. Additionally, a main training split is provided that aggregates data from all these sources. The dataset has undergone preprocessing steps including merging, duplicate removal, and normalization to ensure data uniqueness and consistency.