Synthetic spike-in metabarcoding for plant pathogen diagnostics results in precise quantification of copy number...

Synthetic spike-in metabarcoding (SSIM) assays generate quantitative next-generation sequencing (NGS) data, but are marred by inconsistency and have seen limited adoption. Previous efforts to develop SSIM assays have focused on the ITS and 16S rRNA genes. This study marks the first use of SSIM as a diagnostic assay to identify and quantify plant-pathogenic species within the genus Fusarium and implements it using the single-copy TEF1 gene, which has relatively uniform G + C content and length. We identified variability between species in read quality score as a key source of bias that impacts SSIM to a lesser extent than other quantitative NGS approaches. SSIM was validated against another quantitative NGS assay that utilized qPCR (qMET) to calculate the total gene copy number. The comparison showed that SSIM was both precise (R² > 0.93 for three Fusarium species) and proportional (slope ~1) in relation to qMET. Further, we applied SSIM to 24 wheat grain samples from Italy, revealing a diverse array of Fusarium species and associated mycotoxins, with SSIM demonstrating superior predictive accuracy for most toxin concentrations compared to qPCR. Our results underscore the utility of SSIM for pathogen-agnostic diagnostics, with important implications for food safety and management of mycotoxin contamination.