BioThings Explorer

BioThings Explorer queries federated biomedical knowledge graphs to aggregate semantically annotated data from multiple web services and execute chained, multi-step graph queries for analyses such as drug repurposing and target identification.

Key Features:

• Federated Knowledge Graph Structure: Operates on a federated model that integrates diverse biomedical data from multiple web services into a virtual knowledge graph.
• Semantic Precision in Data Annotation: Uses semantically precise annotations for inputs and outputs across resources to ensure accurate interpretation and interoperability.
• Automated Multi-step Query Execution: Automates the chaining of web service calls to perform complex, multi-step graph queries across distributed resources.
• Dynamic Information Retrieval: Retrieves data at query time from underlying web services so results reflect the current state of source data rather than a centralized static repository.

Scientific Applications:

• Drug Repurposing: Enables identification of potential new uses for existing drugs by analyzing interconnected entities and relationships across the knowledge graph.
• Identification of Drug Targets: Supports pinpointing novel therapeutic targets by traversing pathways and associations integrated from multiple data sources.
• Prediction of Drug Side Effects: Facilitates prediction of adverse effects by linking compound, gene, and phenotype information across federated resources.
• Clinical Decision Support: Provides integrated biomedical evidence from multiple sources to inform clinical decision-making and treatment strategies.

Methodology:

Aggregates data from multiple disparate web-service sources into a virtual federated knowledge graph, uses semantically annotated inputs and outputs to enable integration, and executes automated chaining of web service calls to perform complex multi-step graph queries.