A Structure-Derived Framework for Interpreting Membrane Permeability Beyond logP and Polar Surface Area

Membrane permeability is a key determinant of oral drug absorption and underlies the Biopharmaceutics Classification System (BCS), yet its physicochemical basis remains difficult to interpret using conventional in silico descriptors. Although lipophilicity and polar surface area capture broad trends, they do not fully explain permeability differences among compounds with similar global properties. Here we present a structure-derived framework that reorganizes commonly available physicochemical descriptors into a unified multidimensional representation without relying on experimental permeability or solubility data. Within this space, permeabilityrelated characteristics emerge as continuous spatial relationships rather than predefined categories. Compounds associated with BCS Classes II and IV occupy contiguous but distinguishable regions, while compounds with ambiguous classifications tend to localize near transitional boundaries. The observed organization reflects differences in the relative balance of dispersive, polar, and hydrogen-bonding interactions that govern desolvation and membrane crossing. Rather than redefining regulatory classification schemes, the proposed framework provides a structure-based perspective for interpreting permeability-related behavior, particularly in early stages of drug discovery where experimental data are limited.