Computationally Guided Cannabinoid Discovery for Antiviral and Anticancer Applications: Chemical Space, Molecular Modeling, and Artificial Intelligence

Cannabinoids represent a structurally diverse class of molecules with potential applications in antiviral and anticancer drug discovery. However, their conformational flexibility, polypharmacology, physicochemical limitations, and heterogeneous biological effects complicate the identification of therapeutically relevant candidates. This review examines how computer-aided drug discovery, molecular simulation, quantum chemistry, quantitative structure–activity relationship modelling, and artificial intelligence can be integrated to investigate cannabinoid chemical space and prioritize candidate compounds. Antiviral applications, including SARS-CoV-2 proteases and viral-entry mechanisms, are considered alongside anticancer investigations involving receptor tyrosine kinases, cannabinoid-sensitive receptors, and intracellular signalling pathways. Particular emphasis is placed on distinguishing computational predictions from experimentally demonstrated molecular binding, target inhibition, and cellular activity. The complementary contributions and limitations of molecular docking, molecular dynamics, binding free-energy estimation, and ADMET prediction are examined in relation to candidate prioritization. The review further discusses dataset quality, model validation, uncertainty estimation, and the challenges of translating predicted interactions into biologically meaningful outcomes. An integrated computational–experimental framework is proposed in which candidate selection, biochemical and biophysical validation, cellular evaluation, selectivity assessment, and structure–activity relationship analysis inform iterative lead optimization. Future progress will depend on experimentally anchored AI-assisted design, improved uncertainty quantification, and explicit consideration of pharmacokinetics, toxicity, and target-specific biological context. This framework provides a basis for more rigorous evaluation of cannabinoid-derived candidates while recognizing that most antiviral and direct anticancer applications remain at the preclinical or investigational stage.