EVisionary: A Harmonized Multi-Omics Extracellular Vesicle Database (v3.2)

The EVisionary Framework (v3.2) provides a rigorously harmonized, multi-omics Extracellular Vesicle (EV) metadata repository. This dataset was created to resolve the pervasive metadata fragmentation, semantic inconsistencies, and structural conflations present in legacy EV databases (e.g., Vesiclepedia, ExoCarta, EV-TRACK). Data Structure & Contents The unified database contains 238,891 curated EV records, integrating cross-kingdom species data, molecular identifiers (Proteins, RNAs, Lipids, Metabolites), and isolation method metadata. To ensure programmatic scalability and privacy-preserving local execution, the dataset is provided in a highly compressed columnar Apache Parquet format (unified_EVmetadata.parquet), optimized for DuckDB and Pandas analytics. Key Enhancements Semantic Normalization: Elimination of case-sensitivity issues and duplicate identifier conflations. Multi-Omics Integration: Seamless querying across diverse molecular types. MISEV Compliance Tracking: Foundation for contextualized biomarker discovery. Usage This dataset serves as the analytical foundation for the article: “EVisionary: A scalable integration framework and quantitative audit for harmonizing extracellular vesicle repositories”. For reproducible scripts, text-mining pipelines, and API utilization, please refer to the linked GitHub repository.