Systems pharmacology investigation of mechanism of action of nutraceuticals

Abstract Nutraceuticals’ properties render them a good source of potential drug leads in pharmaceutical research, but the elucidation of their mechanism of action (MoA) is challenging, partly because natural compounds can bind multiple targets with unrelated structures. In this work, we present a pipeline that integrates complementary systems biology and chemoinformatics-based approaches with the goal of identifying potential targets and generating MoA hypotheses that can be tested in a new cycle of experiments. We focus on how we can integrate knowledge from multiple sources into an ensemble score and translate our results into biological insights, providing use cases on three extensively studied nutraceuticals, epigallocatechin-3-gallate, quercetin, and resveratrol as well as on a prototypical proinflammatory stimulus, TNF-α. Fifty-seven of our top 60 predicted targets were already identified in the literature as experimentally perturbed by the corresponding compounds, interestingly observing anticancer, antiviral, and antibacterial MoA to be shared among the tested nutraceuticals.