In drug discovery, machine-learning models can help prioritize promising target–disease hypotheses, but their predictions are often difficult to interpret or act on without clear supporting evidence. In this webinar, attendees will learn how knowledge graphs can help turn prediction scores into explainable, evidence-backed hypotheses by connecting them to relevant mechanisms, pathways, biomarkers, drugs, safety signals, and clinical context.
We’ll showcase how Clarivate’s knowledge graph, its AI-powered natural-language interface and integration with trusted content from CDDI, MetaBase, and OFF-X which creates a connected evidence network spanning 2.4 million nodes and over 10 million edges can support the interpretation of machine-learning predictions. GLP1R drug repurposing will be the practical case study, illustrating how scientists can investigate why specific GLP1R–disease pairs receive high prioritization scores, retrieve supporting evidence, and build a coherent biological rationale for predicted associations.
You’ll also see how AI-powered knowledge graphs can make machine-learning predictions more transparent, interpretable, and actionable across scientific and cross-functional teams and learn the answers to questions such as:
- How can knowledge graphs and natural-language interfaces help make machine-learning predictions easier to interpret?
- How can Clarivate’s biomedical knowledge graph support the investigation of GLP1R indication-prioritization hypotheses?
- How can evidence from targets, pathways, biomarkers, drugs, safety signals, and clinical context be connected into a coherent biological rationale?
