Enhancing Rare Disease Research with Semantic Integration of Environmental and Health Data

Knowledge Graph (KG) approaches are increasingly being used for data integration processes to combine clinical data with other data sources. Health Data Researchers (HDR) could benefit from these technologies since they require additional types of data outside the health sector, like environmental data, to better understand the extrinsic factors that influence health outcomes in rare disease research. However, using and directly navigating the combined data in the KG can be an obstacle for HDRs. To address this problem, the Semantic Environmental and Rare Disease data Integration Framework (SERDIF) was designed to hide the complexities for these researchers when exploring linked environmental observations with clinical data using a KG approach. The framework was evaluated by HDRs for a case study on Anti-neutrophil cytoplasm antibody (ANCA)-associated vasculitis (AAV) in Ireland, and promising usability and effectiveness results were observed. HDRs studying AAV were able to access, explore and export environmental related data to be used as input for their statistical models. SERDIF has the potential to be a solution for HDRs, who require a flexible methodology to integrate environmental data with longitudinal and geospatial diverse clinical data, in their hypothesis validation of environmental factors for rare disease research.