PrimeKG

Here, we present the Precision Medicine Knowledge Graph (PrimeKG). This resource provides a holistic view of diseases. We have integrated 20 high-quality datasets, biorepositories and ontologies to curate this knowledge graph. PrimeKG systematically captures information about 17,080 diseases with 4,050,249 relationships representing various major biological scales, including diseases, drugs, genes, proteins, exposures, phenotypes, drug side effects, molecular functions, cellular components, biological processes, anatomical regions, and pathways. Disease nodes in our multi-relational knowledge graph are densely connected to every other node type. PrimeKG's rich graph structure is supplemented with textual descriptions of clinical guidelines for drug and disease nodes to enable multi-modal disease exploration. To get started with using PrimeKG, please explore our project website: https://zitniklab.hms.harvard.edu/projects/PrimeKG/