Etablierung der § 64 LFGB Arbeitsgruppen „NGS – Bakteriencharakterisierung“ und „NGS – Speziesidentifizierung“

Benchmarking and Validation of a Bioinformatics Workflow for Meat Species Identification Using 16S rDNA Metabarcoding

DNA-metabarcoding is becoming more widely used for routine authentication of meat-based food and feed products. Several methods validating species identification methods through amplicon sequencing have already been published. These use a variety of barcodes and analysis workflows, however, no methodical comparison of available algorithms and parameter optimization are published hitherto for meat-based products' authenticity. Additionally, many published methods use very small subsets of the available reference sequences, thereby limiting the potential of the analysis and leading to over-optimistic performance estimates. We here predict and compare the ability of published barcodes to distinguish taxa in the BLAST NT database. We then use a dataset of 79 reference samples, spanning 32 taxa, to benchmark and optimize a metabarcoding analysis workflow for 16S rDNA Illumina sequencing. Furthermore, we provide recommendations as to the parameter choices, sequencing depth, and thresholds that should be used to analyze meat metabarcoding sequencing experiments. The analysis workflow is publicly available, and includes ready-to-use tools for validation and benchmarking.

A next-generation sequencing approach for the detection of mixed species in canned tuna

Tuna cans are relevant seafood products for which mixtures of different tuna species are not allowed according to European regulations. In order to support the prevention of food fraud and mislabelling, a next-generation sequencing methodology based on mitochondrial cytochrome b and control region markers has been tested. Analyses of defined mixtures of DNA, fresh tissue and canned tissue revealed a qualitative and, to some extent, semiquantitative identification of tuna species. While the choice of the bioinformatic pipeline had no influence in the results (p = 0.71), quantitative differences occurred depending on the treatment of the sample, marker, species, and mixture (p < 0.01). The results revealed that matrix-specific calibrators or normalization models should also be used in NGS. The method represents an important step towards a semiquantitative method for routine control of this analytically challenging food matrix. Tests of commercial samples uncovered mixed species in some cans, being not in compliance with EU regulations.

Food authentication in real life: How to link nontargeted approaches with routine analytics?

In times of increasing globalization and the resulting complexity of trade flows, securing food quality is an increasing challenge. The development of analytical methods for checking the integrity and, thus, the safety of food is one of the central questions for actors from science, politics, and industry. Targeted methods, for the detection of a few selected analytes, still play the most important role in routine analysis. In the past 5 years, nontargeted methods that do not aim at individual analytes but on analyte profiles that are as comprehensive as possible have increasingly come into focus. Instead of investigating individual chemical structures, data patterns are collected, evaluated and, depending on the problem, fed into databases that can be used for further nontargeted approaches. Alternatively, individual markers can be extracted and transferred to targeted methods. Such an approach requires (i) the availability of authentic reference material, (ii) the corresponding high-resolution laboratory infrastructure, and (iii) extensive expertise in processing and storing very large amounts of data. Probably due to the requirements mentioned above, only a few methods have really established themselves in routine analysis. This review article focuses on the establishment of nontargeted methods in routine laboratories. Challenges are summarized and possible solutions are presented.