Models for Nonspecific Binding and Partitioning
Nonspecific binding or partitioning plays a major role in the distribution and disposition of drugs. Plasma protein binding and partitioning into tissue membranes are two important processes that influence unbound drug concentrations and therefore most pharmacological and toxicological properties. Multiple binding proteins and multiple binding sites on these proteins complicate models for plasma protein binding. Computational models for plasma protein binding can generally explain ∼70% of the variance for diverse datasets. Nonspecific binding processes are important when scaling up microsomal or cellular data to predict clearance and drug interactions, and models for microsomal membrane binding and nonspecific binding to hepatocytes are available. Distribution into cells and organelles due to pH differences can be readily modeled by pH partitioning equations. Models for the volume of distribution include descriptor-based models and mechanistic models. Within the physiologically based pharmacokinetic framework there are integrated models to predict volume of distribution that incorporate plasma protein binding, lipid partitioning, and pH partitioning. However, the current implementations of these models require some experimental data.