In silico approach to safety of botanical dietary supplement ingredients utilizing constituent-level characterization

Abstract Botanicals used in dietary supplements industry can have toxicology concerns related to endpoint gaps that cannot be fully resolved by a history of use, or existence of conflicting safety data. However, traditional toxicological studies on botanicals are scientifically and pragmatically challenging due to testing of complex mixtures of constituents, cost, time, and animal usage. Alternatively, we developed an in silico decision-tree approach to address data gaps and inform need for further studies by toxicologically evaluating the chemical composition of botanicals. Following advanced multi-detector analytical characterization of a botanical, each chemical constituent is: (a.) quantitatively benchmarked against similar constituents in commonly consumed foods or botanicals with well-established safety profiles, (b.) systematically evaluated for toxicity data utilizing structure-activity relationships, and, (c.) compared to established thresholds of toxicological concern in absence of safety data or structural analogs. Finally, where safety endpoint gaps are identified which cannot be resolved without additional in vitro or in vivo studies, the botanical compositional data are critical to inform on study design. Results with three herbal preparations demonstrate the utility of this novel approach to identify potential hazards and establish safe human use levels for botanicals in a cost efficient and informative manner that minimizes animal use.