Estimating human ADME properties, pharmacokinetic parameters and likely clinical dose in drug discovery

<b>Introduction</b>: Prediction of human absorption, distribution, metabolism, and excretion (ADME) properties, therapeutic dose and exposure has become an integral part of compound optimization in discovery. Incorporation of drug metabolism and pharmacokinetics into discovery projects has largely tempered historical drug failure due to sub-optimal ADME. In the current era, inadequate safety and efficacy are leading culprits for attrition; both of which are dependent upon drug exposure. Therefore, prediction of human pharmacokinetics (PK) and dose are core components of de-risking strategies in discovery. <b>Areas covered</b>: The authors provide an overview of human dose prediction methods and present a toolbox of PK parameter prediction models with a proposed framework for a consensus approach valid throughout the discovery value chain. Mechanistic considerations and indicators for their application are discussed which may impact the dose prediction approach. Examples are provided to illustrate how implementation of the proposed strategy throughout discovery can assist project progression. <b>Expert opinion</b>: Anticipation of human ADME, therapeutic dose and exposure must be deliberated throughout drug discovery from virtual/initial synthesis where key properties are considered and similar molecules ranked, into development where advanced compounds can be subject to prediction with greater mechanistic understanding and data-driven model selection.