BacPath: A Rapid Metagenomic Workflow with a Curated Reference Database for Species-Level Detection of Bacterial Plant Pathogens

The emergence of plant pathogens through geographic spread, host shifts, and climate change threatens disease management and global food security. Detection of emerging pathogens without prior genomic information is critical for preventing widespread outbreaks and enabling timely management to reduce economic impact. Accurate and reliable detection of infecting organisms using metagenome-based bioinformatic pipelines is therefore essential. To address this, we have developed and validated BacPath pipeline and a highly accurate complete genome database, BacPathDB. Raw read files obtained from Oxford Nanopore sequencers (MinIon and Flongle) were processed, quality-scored, and mapped against a curated database. The expandable BacPathDB contains high-quality genomes with accurate metadata and currently includes 317 genomes, representing species, within the genera Dickeya and Pectobacterium, including most type strains. The pipeline accurately detected all 32 species within these two genera in silico, with no false positives or false negatives. In vitro analyses with both Pectobacterium and Dickeya species further confirmed the accuracy of the developed pipeline for detecting pathogens from complex plant tissue samples infected with single or multiple pathogens. Detection can be achieved with as little as a single read, however, analyses based on five or more reads provide higher confidence in the results. The pipeline is rapid, enabling detection in approximately one minute, and can be performed on a 16 GB laptop without requiring high computational resources. This makes it well suited for field applications, while the curated reference database enhances its reliability for routine diagnostics, epidemiological investigations, disease surveillance, regulatory programs, biosecurity, and microbial forensics.