APObind

APObind是由国际信息技术研究所创建的数据集，专注于提供PDBbind数据集中蛋白质的无配体结合状态（apo）构象。该数据集包含10,599个无配体蛋白质结构，旨在解决现有机器学习模型在处理无配体蛋白质结构时的局限性。创建过程中，通过序列比对和结构分析确保了apo结构的准确性。APObind的应用领域主要集中在药物设计中，特别是在预测蛋白质结合位点和结合亲和力方面，为机器学习模型提供了更全面的训练和验证数据。

APObind is a dataset developed by the International Institute of Information Technology, which focuses on providing the ligand-free (apo) conformations of proteins from the PDBbind dataset. This dataset contains 10,599 ligand-free protein structures, aiming to address the limitations of current machine learning models when processing ligand-free protein structures. During its construction, the accuracy of these apo structures was verified via sequence alignment and structural analysis. APObind is primarily applied in drug design, especially in protein binding site prediction and binding affinity prediction, providing more comprehensive training and validation data for machine learning models.