Why Graph?

Biology-inspired data model to faithfully capture biological complexity.

Biology is inherently complex and difficult to model. Physics and chemistry have well-defined laws and equations (e.g., Newton's laws, thermodynamics, and quantum mechanics) that provide a high degree of predictability and precision in modeling phenomena. These frameworks make it possible to calculate specific outcomes, like the exact trajectory of an asteroid thousands of lightyears away, with extreme precision.

Biology deals with emergent properties—features that materialize only when complex networks of interacting molecules, cells, and organisms come together, and which cannot be easily inferred by studying each component in isolation. This complexity is especially pivotal in drug discovery, where the old paradigm of investigating a single target in a vacuum often fails to predict how a compound will behave in a living system.

So then, why graph? Because the reality is that biology doesn’t conform neatly to the confines of traditional tabular data structures. To fully capture these emergent phenomena, researchers increasingly turn to graph-based approaches—knowledge graphs, for instance—that model the rich, contextual interplay among genes, proteins, metabolic pathways, and cellular processes. By mapping data onto interconnected graphs rather than flattening it into tables, drug developers can more effectively trace subtle influence patterns, anticipate off-target effects, and design interventions that align with the full complexity of biological systems. In this network-driven environment, emergent properties become navigable landscapes of relationships, enabling a more holistic and predictive framework for drug discovery, ultimately guiding the design of therapies better attuned to the intricacies of biological complexity.