Data for: Assessment of Several Machine Learning Methods Towards Reliable Prediction of Hormone Receptor Binding Affinity

Experimental binding affinities (logRBA) were retrieved from the Estrogenic Activity Database (EADB) and compounds with large uncertainties were removed to yield 1589 data points. Molecular descriptors were computed using 3D molecular structures that were generated using their respective SMILES codes in EADB. The data may be used to train and evaluate the performance of machine learning methods or binding free energy calculations.实验结合亲和力（logRBA）数据从雌激素活性数据库（Estrogenic Activity Database，EADB）中获取，随后剔除了不确定性较高的化合物，最终得到1589组有效数据点。 分子描述符的计算基于3D分子结构，而这些3D分子结构是通过该数据库中各化合物对应的SMILES（Simplified Molecular Input Line Entry System）编码生成的。本数据集可用于训练机器学习方法并评估其性能，亦可用于结合自由能计算方法的性能验证。