VTAM: A robust pipeline for validating metabarcoding data using internal controls

Abstract Metabarcoding studies should be carefully designed to minimize false positives and false negative occurrences. The use of internal controls, replicates, and several overlapping markers is expected to improve the bioinformatics data analysis. VTAM is a tool to perform all steps of data curation from raw fastq data to taxonomically assigned ASV (Amplicon Sequence Variant or simply variant) table. It addresses all known technical error types and includes other features rarely present in existing pipelines for validating metabarcoding data: Filtering parameters are obtained from internal control samples; cross-sample contamination and tag-jump are controlled; technical replicates are used to ensure repeatability; it handles data obtained from several overlapping markers. Two datasets were analysed by VTAM and the results were compared to those obtained with a pipeline based on DADA2. The false positive occurrences in samples were considerably higher when curated by DADA2, which is likely due to the lack of control for tag-jump and cross-sample contamination. VTAM is a robust tool to validate metabarcoding data and improve traceability, reproducibility, and comparability between runs and datasets.