Spatial Omics Rewrites the Computational Map of Drug Exposure, Target Engagement, Response, and Resistance. Download PDF

Pharmaceutical models commonly represent drug action through measurements aggregated across tissues, specimens, or patient populations. These averages can obscure the spatial separation between drug delivery, molecular target engagement, cellular response, adaptive resistance, and tissue toxicity. Spatial omics creates an opportunity to reconsider these relationships by preserving the locations of molecular states, cell populations, histologic structures, and pharmacologic signals. This Perspective develops a proposed Tissue-Aware Pharmacologic State Map that organizes drug action as four linked but non-equivalent spatial fields: local exposure, local target engagement, neighborhood-conditioned response, and resistance or toxicity. The framework treats tissue neighborhoods, anatomical barriers, vascular structures, stromal compartments, immune organization, and local metabolic states as components of the pharmacologic context rather than as secondary annotations. It also distinguishes directly measured spatial evidence from computationally reconstructed or predicted information. Under this formulation, a compound detected within tissue does not necessarily reach its intended molecular target; target expression does not establish target engagement; engagement does not guarantee a therapeutically meaningful response; and localized response does not establish organism-level efficacy or safety. The proposed construct therefore rejects the use of a single predictive or benchmarking measure as sufficient evidence of pharmaceutical usefulness. Instead, claims must be evaluated across assay validity, spatial registration, computational representation, perturbational consistency, external transportability, uncertainty, and decision-specific utility. The framework remains conceptual and requires empirical evaluation across drug classes, tissues, spatial platforms, and translational settings. Its principal implication is that computational pharmacology should model where drug action occurs, under which tissue conditions it occurs, and at which spatial and evidentiary scale a claim remains valid.