Mapping the off-target effects of every FDA-approved drug in existence (www.owlposting.com)

EvE Bio, a nonprofit Focused Research Organization, has produced and released a comprehensive dataset mapping off‑target interactions between a broad panel of human cellular receptors and roughly 1,600 FDA‑approved drugs. The resource is published under a non‑commercial Creative Commons-style license (free for academics, commercially licensable) and represents a pre‑triaged matrix of drug–target binding/perturbation data intended to reveal unintended receptor interactions across a large clinically relevant compound set. This matters for AI/ML and drug discovery because it provides high‑value ground truth for three near‑term use cases: systematic drug repurposing (shortening development time by leveraging known safety profiles), rigorous validation/benchmarking of target‑prediction and toxicity models, and deeper polypharmacology analyses. Technically, the dataset enables supervised learning and calibration of predictive models, supports mechanistic hypothesis generation for off‑label effects, and can prioritize experimental follow‑ups—while also carrying practical caveats (experimental validation still required, economic barriers to commercial repurposing, and license limits on for‑profit use). By freeing academic access to large, standardized off‑target maps, EvE’s work should substantially improve model training, reduce blind spots in safety/ToxAI, and accelerate identification of clinically actionable repurposing leads.