Biomedical knowledge graphs now integrate millions of entities and relationships across genes, diseases, drugs, pathways, phenotypes, and experimental evidence. But for target discovery, the value of a graph depends less on how many connections it contains and more on whether those connections reflect the underlying biology.

Widely used resources such as Open Targets make it easier to aggregate and rank gene-disease evidence. But many off-the-shelf knowledge graphs flatten different kinds of evidence into broad associations, losing the directionality, evidence type, certainty, and biological context behind each relationship. A gene-disease link, for example, could mean that a genetic variant increases disease risk, loss of gene function is protective, expression changes downstream of pathology, or experimental perturbation shows that the gene drives progression. These should not become equivalent edges in a knowledge graph.

In this webinar, we'll show how models can recover these distinctions from biomedical literature and convert them into structured, traceable relationships: separating causal claims from molecular observations, distinguishing biomarkers from therapeutic targets, and capturing whether an effect promotes or protects against disease.

Using acute myeloid leukemia (AML) as a case study, we'll show how typed relationships make evidence reasoning across a knowledge graph more reliable.