Table 4.

List of information in the final datasets.

Column Name	Description	Data Type
Smiles_unify	Standardized SMILES representation of compounds is based on standardization methods described in the Data Standardization.	String
value	Experimental values for different datasets including regression values and classification values	Float
property	Different ADMET property name for the experiment	String
scaffold_train_test_label	Training labels based on scaffold splitting, where 1 represents the training data and 0 represents the testing data.	Float
random_train_test_label	Training labels based on random splitting, where 1 represents the training data and 0 represents the testing data.	Float

Including the column name, the description for the column information and the data type.