CataAI/PoseX

该数据集包含三个子数据集：Astex Diverse set、PoseX Self-Docking set和PoseX Cross-Docking set。Astex Diverse set由Hartshorn等人整理，包含85个案例，已重新处理以适应评估过程。PoseX Self-Docking set和PoseX Cross-Docking set由我们使用PDB数据库整理，分别包含718个案例和1312对案例。每个蛋白质-配体复合物都有一个以PDB标识符和CCD标识符命名的文件夹，包含蛋白质结构文件、配体文件、分子构象文件、JSON输入文件以及交叉对接集特有的参考蛋白质结构和组信息文件。此外，还包括计算口袋结构相似性的CSV文件。

The dataset consists of three sub-datasets: Astex Diverse set, PoseX Self-Docking set, and PoseX Cross-Docking set. The Astex Diverse set, curated by Hartshorn et al., contains 85 cases and has been reprocessed to fit our evaluation process. The PoseX Self-Docking set and PoseX Cross-Docking set, curated using the PDB database by us, contain 718 cases and 1312 pairs respectively. Each protein-ligand complex has a folder named after the PDB identifier and the CCD identifier, containing protein structure files, ligand files, molecular conformation files, JSON input files, and reference protein structure and group information files unique to the Cross-Docking set. Additionally, there are CSV files that contain pocket structure similarity calculations for each evaluation data point relative to the training data.