Systematic Evaluation of Metagenomic Sequencing Workflows for Pathogen Identification and Relative Quantification in Mock Coastal Seawater

Pathogen identification and relative quantification of seawater-borne microbes are essential for safeguarding public health and evaluating marine ecosystem dynamics. However, the performance of metagenomic next-generation sequencing (mNGS) in coastal seawater remains poorly assessed due to matrix complexity and workflow variability. Here, we established seawater-mimicking reference samples containing 14 representative microbes, and systematically evaluated 28 workflows integrating four DNA extraction methods, three sequencing platforms with corresponding library preparation kits, and three taxonomic classifiers. Performance was assessed using precision, recall, and F1-score for qualitative accuracy together with false-positive relative abundance and similarity for quantitative accuracy. Among the tested variables, the DNeasy PowerWater kit yielded the highest DNA recovery efficiency, Illumina NovaSeq 6000 with TrueLib mRNA Library Prep Kit consistently outperformed the others, and Kraken2-Bracken provided the favorable taxonomic assignments under optimizing thresholds conditions. Workflows combining the DNeasy PowerWater or PowerSoil Pro kit, Illumina NovaSeq 6000 with TrueLib mRNA Library Prep Kit and Kraken2-Bracken achieving the best overall accuracy, with genus-level F1-scores exceeding 0.90. In contrast, other combinations, such as the DNeasy PowerSoil Pro kit or CTAB method, MGISEQ-2000 platform with TruePrep Flexible DNA Library Prep Kit and Kaiju, yielded F1-scores below 0.40 under optimizing thresholds conditions. This study highlights the necessity of methodological pre-evaluation to ensure robust data for public health and environmental applications.