BioDataFuse: Enhancing Data Interoperability through Modular Queries and Knowledge Graph Construction

In biological research, integrating experimental data with publicly available resources is pivotal for understanding complex biological mechanisms. However, this process is often intricate and time-consuming due to the complexity and diversity of data. Furthermore, the lack of consistent harmonization across different data types complicates the management of disparate data formats and sources. Addressing this, we introduce BioDataFuse, a query-based Python tool for seamless integration of biomedical data resources. BioDataFuse establishes a modular framework for efficient data wrangling, enabling context-specific knowledge graph creation and supporting graph-based analyses. With a user-friendly interface, it enables users to dynamically create knowledge graphs from their input experimental data. Supported by a robust Python package, pyBiodatafuse, this tool excels in data harmonization, aggregating diverse sources through modular queries. Moreover, BioDataFuse provides plugin capabilities for Cytoscape and Neo4j, allowing local graph hosting. Ongoing refinements enhance the graph utility through tasks like link prediction, making BioDataFuse a versatile solution for efficient and effective biological data integration.