Generation of open biomedical datasets through ontology-driven transformation and integration processes

Area of Science:

• Biomedical Informatics
• Data Science
• Semantic Web Technologies

Background:

• Biomedical research faces challenges in integrating large, heterogeneous data from multiple sources.
• The Semantic Web and Linked Open Data (LOD) offer solutions for machine-readable data integration and sharing.

Purpose of the Study:

• To develop and present an approach for transforming and integrating heterogeneous biomedical data into Semantic Web formats.
• To generate open biomedical datasets adhering to Linked Open Data principles.

Main Methods:

• Data transformation based on mappings between data schemas and ontological infrastructure.
• Utilizing complex transformation patterns and enabling mapping reusability.
• Automated application of mappings to generate logically consistent, machine-readable content.

Main Results:

• A standardized process for transforming and integrating heterogeneous biomedical data.
• Application of LOD principles to create interoperable, open biomedical datasets.
• Development and application of a software tool (SWIT) for data transformation and integration.

Conclusions:

• The presented approach successfully generates open biomedical repositories in Semantic Web formats.
• SWIT enables the application of LOD principles, facilitating the creation of linked, open, and integrated biomedical datasets.
• The method allows for the integration of data from multiple sources and schemas, enhancing data interoperability.