PrimeKG
A biomedical knowledge graph integrating 20 resources to describe 17,080 diseases with over 4 million relationships across ten biological scales.
What is PrimeKG?
PrimeKG (Precision Medicine Knowledge Graph) is a biomedical knowledge graph that integrates data from 20 resources to describe 17,080 diseases with over 4 million relationships across ten biological scales. It connects diseases with genes, drugs, phenotypes, and clinical information to enable data-driven research in precision medicine. The project provides ready-to-use datasets and tools for building and updating the knowledge graph.
Bioinformaticians, medical researchers, and data scientists working on precision medicine, drug discovery, or disease modeling who need integrated biomedical data for analysis.
PrimeKG offers a uniquely comprehensive and clinically relevant knowledge graph that bridges multiple biological scales, with multimodal data and extensive coverage of diseases, including rare conditions. Its integration of diverse resources and ready-to-use format saves researchers time in data preparation.
Overview
Precision Medicine Knowledge Graph (PrimeKG)
Use Cases
Best For
- Integrating heterogeneous biomedical data for precision medicine research
- Building machine learning models for drug discovery or disease prediction
- Analyzing relationships between diseases, genes, and drugs at scale
- Studying rare diseases with comprehensive clinical and molecular data
- Creating knowledge graph embeddings for biomedical entity relationships
- Developing data pipelines for updated biomedical knowledge graphs
Not Ideal For
- Projects requiring real-time or frequently updated biomedical data for dynamic applications
- Teams without dedicated bioinformatics or data engineering resources to manage complex script-based updates
- Applications needing interactive, graph-based query interfaces out-of-the-box without additional development
- Research focused narrowly on a single data source where simpler, specialized datasets would suffice
Pros & Cons
Pros
Includes over 17,000 diseases, including rare ones, optimized for clinical relevance, as highlighted in the README's unique features section.
Augments disease and drug nodes with clinical descriptors from authorities like Mayo Clinic and Orphanet, enabling richer analyses for precision medicine.
Provides pre-processed CSV files for easy download from Harvard Dataverse, saving researchers significant time in data preparation.
Includes scripts to process 20 primary resources and build updated versions, supporting long-term usability as detailed in the July 2023 update.
Cons
Building an updated graph requires running multiple scripts and notebooks, some of which are noted as potentially out-of-date or needing fixes, as seen in the update details.
The knowledge graph is static between releases, which may not suit applications requiring the latest biomedical discoveries or real-time data integration.
Individual datasets have varying licenses, requiring users to navigate legal restrictions for commercial use, as mentioned in the license section.
Frequently Asked Questions
Related Projects
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.