PaVarDB: a database and web platform for missense variant analysis in Pseudomonas aeruginosa using an end-to-end BVbase pipeline.

Genomic variant data are useful in detecting and treating antibiotic-resistant bacteria. However, there are no bacterial genomic variant databases that catalogue the variations in the different genes across strains. In this work a Nextflow- and Docker-based end-to-end pipeline, BVbase, that can automate the creation of databases from raw high-throughput sequences has been created to fill this lacuna with Pseudomonas aeruginosa as a case study. Pseudomonas aeruginosa is a Gram-negative adaptable pathogen with multiple antibiotic resistances that causes various types of infections, including respiratory, urinary, and bloodstream infections. The pipeline can take multistrain genomic files, detect missense variants, and save results in a database with the help of Python and SQLite (https://github.com/bic-sastra/BVbase). Using the generated database for P. aeruginosa, a web application interface has been made using Flask and HTML that runs in a server with MySQL backend (https://bic.sastra.edu/pavardb). The web application provides supports for different types of queries to select variants by gene, geographical group, isolation country, antibiotics, and resistance phenotype. This web interface generates results as variant tables, plots, and statistics for the selected data. By enabling interactive visualizations and advanced selection, the platform supports research and clinical use through the exploration of genomic variations associated with antimicrobial resistance.