Additional file 1 of Ensemble learning for predicting ex vivo human placental barrier permeability

Additional file 1: Table S1. The features, experimental CI, predicted CI values, and AD of the training dataset. Table S2. The mean and variance values of the 7 selected informative features for the ensemble model. Table S3. The features, experimental CI, predicted CI values, and AD of the test dataset. Table S4. The rules for defining the applicability domain of the ensemble model based on the training dataset. Table S5. Details of the final prediction model based on all 87 chemicals. Table S6. The rules for defining the applicability domain of the final prediction model. Table S7. The features, experimental CI, predicted CI values, and AD of the external test dataset. Figure S1. The comparison of experimental and predicted CI values of 87 training dataset for leave-one-out cross-validation. Abbreviations: AD, applicability domain; Y, chemicals within the AD (blue dot); N, chemicals out of the AD (red dot).