Hi Open Targets team,
Open Targets is one of the most valuable biomedical knowledge graph resources available today, integrating genes, diseases, drugs, and multiple evidence types.
I’m curious why there isn’t a production-ready knowledge graph product available (e.g., downloadable graph, Neo4j database, graph API, or KG snapshots) for direct use in graph analytics, machine learning, and AI applications.
Are there ongoing efforts or plans in this direction? It seems like the community could benefit greatly from broader access to the underlying knowledge graph.
Thanks!
Hi Guo,
I’m Paul, and I work with the Open Targets team. My primary responsibility is the construction and maintenance of the Open Targets Platform Knowledge Graph. Although the graph has not yet been officially released, you can download the Neo4j database dump here. The link also includes instructions for importing the database, as well as an interactive meta-graph to help you quickly understand its structure and contents.
https://drive.google.com/drive/folders/17U-uFVmck_chZ9Ap9Ux5_XAf7tC9_i_9?usp=sharing
At the moment, I’m working to host the entire graph on the EMBL-EBI Kubernetes cluster, although we’re still encountering some technical challenges.
My project is funded by Open Targets. One of its goals is to explore AI-assisted knowledge graph maintenance. Depending on the research outcomes and user demand, this work may eventually be adopted by the core team and become part of the official Open Targets Platform. However, it is still too early to say whether that will happen.
I’d be very interested to hear about your use case and receive your feedback. Please feel free to reach out if you run into any issues or would simply like to discuss the project.