KG-Hub

KG-Hub provides standardized construction, exchange, and reuse of biological knowledge graphs to integrate heterogeneous datasets and support semantic inference across biomedical domains.

Key Features:

• Standardized Construction: Employs a modular extract-transform-load pattern and produces knowledge graphs compliant with the Biolink Model.
• Ontology Integration: Supports integration of Open Biological and Biomedical Ontologies (OBO) to incorporate structured vocabularies.
• Data Management: Provides cached downloads of upstream data sources and versioned builds with stable URLs for reproducibility and traceability.
• Cloud Infrastructure: Utilizes cloud infrastructure for storage of KG artifacts.
• Reusability: Enables reuse of transformed subgraphs across projects.

Scientific Applications:

• COVID-19 Research: Facilitates integration and analysis of datasets relevant to SARS-CoV-2 and COVID-19.
• Drug Repurposing: Enables exploration of existing drugs for new indications by linking drug, target, and disease data.
• Microbial-Environmental Interactions: Supports studies of interactions between microbes and their environments through integrated data.
• Rare Disease Research: Assists identification and analysis of rare diseases via comprehensive data integration.

Methodology:

Computational methods include a modular extract-transform-load pattern producing Biolink Model–compliant graphs, integration of OBO ontologies, cached downloads and versioned builds with stable URLs, cloud storage of KG artifacts, and integration with graph machine learning tools for node embeddings, link prediction, and node classification.