Knowledge graph and ontology for disease outbreak scenarios

This project implements a knowledge graph framework for representing disease outbreak scenarios. The knowledge graph is built by processing disease outbreak alerts from ProMED and other sources and combines this with ontological information to create a structured representation of outbreak events.

Sources

The KG builds on the following sources:

• Outbreak alerts: text from outbreak alerts are processed and terms representing diseases/phenotypes/pathogens/symptoms/geolocations are automatically extracted using the Gilda system. This produces mentioned_in relationships.
• Individual alerts are grouped by the inferred outbreak that they belong to, represented as a has_outbreak relationship.
• Taxonomy of diseases: extracted from the Medical Subject Headings tree structure
• Taxonomy of pathogens: extracted from the Medical Subject Headings tree structure
• Taxonomy of geolocations: extracted from the Medical Subject Headings tree structure
• Additional taxonomy of geolocations: extracted from the Geonames dataset with prefix geonames
• Geolocation relations: geolocations are linked together to represent hierarchical inclusion using the isa relationship where the subsumed region is the source of the relationship and the containing region is the target.
• Pathogen-disease relations: relationships representing the fact that a pathogen causes a disease are represented as has_pathogen relationships.
• Disease-phenotype/symptom relations: relationships representing the fact that a disease causes a phenotype/symptom are represented as has_phenotype relationships.
• Development/health indicators: extracted from WDI data, geolocations are linked to indicators with the has_indicator relationship.

Knowledge graph

The knowledge graph is represented as a set of nodes and edges. Nodes represent entities such as diseases, pathogens, geolocations, and phenotypes. Edges represent relationships between these entities. The knowledge graph is deployed in a Neo4j database.

Interaction

The knowledge graph can be queried using the Cypher query language directly through Neo4j. The repository also provides a Python client and a REST API for querying the knowledge graph. The graph database and the surrounding REST API are Dockerized and deployed on AWS.

Funding

This work was supported by CAPTRS.