Bioavailability and ADMET in Drug Discovery

Oral bioavailability and ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) are central determinants of small-molecule drug success, driving decisions from early screening to clinical progression. This paper reviews the conceptual foundations, key physicochemical and biological factors that regulate oral availability, and contemporary in silico strategies for ADMET assessment. It summarizes classical heuristics such as Lipinski’s Rule of Five and Veber’s criteria, explains their predictive strengths and limitations, and describes probabilistic metrics like the Bioavailability Score used in virtual screening. The review outlines major barriers to oral uptake solubility, membrane permeability, first-pass metabolism, and transporter-mediated efflux and connects these barriers to measurable descriptors such as logP, molecular weight, topological polar surface area, rotatable bonds, and hydrogen-bonding capacity. It compares widely used computational tools and web servers (SwissADME, pkCSM, ADMETlab, admetSAR, PreADMET, Molinspiration) in terms of endpoint coverage, methodological basis, and practical application in lead prioritization. Experimental complements including dissolution testing, PAMPA, Caco-2 permeability assays, and in vivo pharmacokinetic measurements are discussed as essential validation steps. The review emphasizes integration of machine learning and large curated datasets to move beyond rigid rule-based filters toward mechanism-informed, data-driven prediction workflows, and highlights the benefit of combining in silico screening with targeted in vitro assays to reduce attrition. The paper concludes that an integrative, iterative approach pairing predictive ADMET modeling with experimental confirmation offers the most efficient path to identify orally bioavailable, safe drug candidates and accelerate translational success.