By Jessica Schneider, Chief Scientific Officer, Corundum Systems Biology

For decades, drug discovery has largely focused on a deceptively simple premise: identify a disease-causing target, design a molecule to modulate it, and translate that effect into clinical benefit. While this reductionist paradigm has delivered important successes, it has also constrained how we conceptualize disease and therapeutic intervention.

Biology is more organized than just single nodes, and increasingly, so are the most impactful interventions on human health. We are now witnessing the maturation and reinterpretation of polypharmacology through the lens of network medicine, enabled by advances in data and artificial intelligence (Hopkins, 2008; Barabási et al., 2011).

Reinterpreting Polypharmacology Through Networks

Polypharmacology has historically been treated as a liability. Off-target effects became synonymous with toxicity while selectivity emerged as a central design principle.

Yet, a retrospective look at many of our most effective therapies suggests a different reality. CNS drugs, kinase inhibitors and numerous small molecules derive their efficacy from engaging multiple targets simultaneously. In place of binary switches, these compounds offer distributed perturbations across biological systems. What was once considered noise is, in many cases, the signal.

Network medicine is now providing the framework for polypharmacology. Biological systems are composed of highly interconnected networks spanning genes, proteins, metabolites, cells, and host–microbiome interactions. Disease occurs following a shift in network state and configuration of interactions that collectively produce pathology and thus disease. Molecular components linked in disease-relevant networks tend to co-organize into “disease modules,” providing a mechanistic basis for multi-target intervention (Barabási et al., 2011).

Importantly, this principle is not unique to pharmaceuticals. Food represents the most pervasive and evolutionarily entrenched form of polypharmacology. Foods are complex mixtures of lipids, micronutrients, metabolites, phytochemicals, peptides, and microbial substrates that act simultaneously across metabolic, immune, endocrine, and microbial pathways. Dietary bioactives such as polyphenols and fibers exert multiple, unrelated effects across signaling cascades, microbiomes and systemic metabolism, demonstrating that coordinated, low-intensity network perturbations drive health effects.

Polypharmacology in Action

Today, polypharmacology is most clearly realized in the emerging class of multi-receptor incretin agonists, which are designed to engage multiple metabolic pathways simultaneously. Building on the success of GLP-1 receptor agonists, dual and triple agonists leverage complementary mechanisms driving greater efficacy than single-target agents. For example, tirzepatide, a GLP-1/GIP dual agonist, has demonstrated superior glycemic control and weight loss compared to selective GLP-1 agonists, while also improving cardiometabolic risk factors (Frias et al., 2021; Jastreboff et al., 2022).

Mechanistically, these agents coordinately modulate distinct but convergent hormonal axes, affecting insulin secretion, appetite regulation, energy expenditure and lipid metabolism in combination. This multi-target engagement represents an evolution to engineered polypharmacology, enabling an irreplicable and potentially synergistic control of complex metabolic disease.

A parallel transformation is occurring in biologics, where bispecific and multispecific antibodies make polypharmacology explicit at the molecular level. , Bispecific antibodies are engineered to simultaneously engage two distinct targets with defined stoichiometry and spatial constraints. In oncology, this has enabled approaches such as T cell engagers that bridge immune effector cells and tumor antigens, effectively rewiring cellular interaction networks rather than modulating a single pathway.

More broadly, bispecific antibodies illustrate that polypharmacology extends beyond promiscuity in binding and can act as a precisely encoded design feature. This can enable coordinated modulation of signaling axes that would otherwise be difficult to influence in concert (Labrijn et al., 2019). Together with incretin-based therapies, these approaches signal a shift: the most effective interventions increasingly act on the relationships within biological systems rather than isolated targets. The success of these drug classes suggests that efficacy may rely on optimizing the targets we are hitting to cover as many bases as possible. This represents a shift away from traditional approaches that attribute efficacy to the strength of drug pharmacology.

The same lens creates a useful bridge between pharmacology and nutrition. Drugs and foods differ in dose, specificity, and regulatory framing. But both can be understood as inputs into biological networks. Drugs tend to deliver higher-amplitude and more targeted perturbations. Food delivers lower-amplitude, distributed signals over longer timescales. As our models improve, the boundary between the two may become less mechanistically distinct than we once assumed.

From Targets to Networks

The convergence of high-dimensional biology and AI has transformed our operational ability, allowing us to move beyond retrospective appreciation of polypharmacology toward prospective design.

A network-based approach can help us move from cataloging molecular parts to modeling state transitions in living systems. This transition will require a corresponding shift in measurement, from static, target-level assays toward rich perturbational datasets that capture how networks respond across time, context, and scale (Hopkins, 2008).

Ultimately, the resurgence of polypharmacology reflects a deeper transition in biology: from reductionism to relational understanding. Interventions used to be defined by the precision of their targets. In the AI era, what matters more is how they reshape the networks that underlie human health.

References

1. Barabási AL, Gulbahce N, Loscalzo J. Network medicine: a network-based approach to human disease. Nature Reviews Genetics. 2011.
2. Hopkins AL. Network pharmacology: the next paradigm in drug discovery. Nature Chemical Biology. 2008.
3. Lincoff AM et al. Semaglutide and cardiovascular outcomes in obesity without diabetes. New England Journal of Medicine. 2023.
4. Ussher JR, Drucker DJ. Glucagon-like peptide-1 receptor agonists: cardiovascular actions and therapeutic potential. Nature Reviews Cardiology. 2023.
5. Labrijn AF, Janmaat ML, Reichert JM, Parren PWHI. Bispecific antibodies: a mechanistic review of the pipeline. Nature Reviews Drug Discovery. 2019.