16S rRNA metabarcoding applied to the microbiome of insect products (novel food): a comparative analysis of three reference databases.

Abstract

The 16S rRNA metabarcoding, based on Next-Generation Sequencing (NGS), is used to assess microbial biodiversity in various matrices, including food. The process involves a "dry-lab" phase where NGS data are processed through bioinformatic pipelines, which finally rely on taxonomic unit assignment against reference databases to assign them at order, genus, and species levels. Today, several public genomic reference databases are available for the taxonomic assignment of the 16S rRNA sequences. In this study, 42 insect-based food products were chosen as food models to find out how reference database choice could affect the microbiome results in food matrices. At the same time, this study aims to evaluate the most suitable reference database to assess the microbial composition of these still poorly investigated products. The V3-V4 region was sequenced by Illumina technology, and the R package "DADA2" was used for the bioinformatic analysis. After a bibliographic search, three public databases (SILVA, RDP, NCBI RefSeq) were compared based on amplicon sequence variant (ASV) assignment percentages at different taxonomic levels and diversity indices. SILVA assigned a significantly higher percentage of ASVs to the family and genus levels compared to RefSeq and RDP. However, no significant differences were noted in microbial composition between the databases according to α and β diversity results. A total of 121 genera were identified, with 56.2% detected by all three databases, though some taxa were identified only by one or two. The study highlights the importance of using updated reference databases for accurate microbiome characterization, contributing to the optimization of metabarcoding data analysis in food microbiota studies, including novel foods.