Phage Bacteria Interaction Pipeline - Data and Test Notebooks

Data Content This dataset provides a curated and ready-to-use database of phage–bacteria interactions, built by integrating multiple public resources such as INPHARED, PhageScope, VRHdb, and related datasets. It focuses on structured metadata and functional annotations for bacteriophages and their bacterial hosts, enabling immediate analysis without additional preprocessing. Detailed information are available in the README file. The software is accessible on GitHub The zenodo software release is accessible here: 10.5281/zenodo.18961927 This data release includes a DuckDB database organized in a star-schema format, along with accompanying CSV files for flexible use. It contains detailed information on phage genomes, host organisms, and predicted functional elements such as proteins, transcription terminators, anti-CRISPR systems, virulence factors, antimicrobial resistance genes, CRISPR arrays, transmembrane proteins, and tRNA/tmRNA features. It also provides curated links between phages and their potential hosts. In addition to the data, the release includes a Jupyter notebook that demonstrates how to explore the database, perform queries, and visualize key properties such as genome characteristics, annotation coverage, and host distributions. Data quality and validation reports are provided as interactive HTML files, documenting dataset consistency, completeness, and integrity checks. Full data accessibility Only metadata and annotations are included in this release. Nucleotide and protein sequences are not provided due to size limitations. Users requiring full sequence data or advanced workflows (e.g., sequence retrieval or machine learning pipelines) should refer to the complete PBI project available on GitHub. The dataset is distributed in a lightweight and accessible format designed for direct use in data analysis, computational biology, and machine learning applications.