Human blood plasma binding
ADMET Predictor™ predicts the percentage of drug that is free in human blood plasma with its 2D S+PrUnbnd model. The model is an artificial neural network ensemble trained with experimental data gathered from the literature for 689 well-characterized molecules. It was tested with a set of 102 molecules that were not applied to the training process in any way.
ADMET Predictor S+PrUnbnd Model Validation
Blood-to-plasma concentration ratio
Many of drug candidates lack reported blood-to-plasma concentration ratio (RBP) values, one of the fundamental pharmacokinetic parameters. RBP is used in both clinical pharmacokinetics and new drug design. ADMET Predictor's S+RBP model attempts to predict this ratio based on 2D structural descriptors. The modeling data gathered from literature consist of 112 compounds, from which 14 were used as a test set.
ADMET Predictor S+RBP Model Validation