CORDIS

Executive Summary:
The European Decathlon (Development of Cost efficient Advanced DNA-based methods for specific Traceability issues and High Level On-site applicatioNsproject) was started on December 1, 2013. The project has finished on November 30, 2016.
The project has focused on the development of DNA-based methods for the three areas of the project: 1) food pathogens, 2) GMOs (genetically modified organisms, with a focus on unauthorised GMOs), and 3) customs issues (with a focus on tobacco and endangered species identification). The work in the different Work Packages within the Decathlon project will be summarized here.

In the area of food pathogens optimised modules were developed for the extraction of DNA directly from larger volumes of food matrices (minced meat or vegetables). Also, two bioinformatics workflows for genome marker identification from larger sets of bacterial genome sequences were developed. Based on the comparative genomic analyses, 47 new specific biomarkers for the clinically most relevant STEC subtypes (O157 and non-O157 ‘big six’ serotypes as well as for O104). A multiplex PCR was developed based on the newly discriminative biomarkers, which allow improved STEC detection and discrimination of the clinically most relevant STEC serotypes. A disc device has been developed that may be used for on-site detection of these markers.

In the field of GMO detection and identification improved methods for GMO detection and identification, including unauthorised GMOs, have been developed. A ddPCR protocol for the identification and quantification of all EU-authorised GM maize varieties has been developed and prevalidated. A method for NGS analysis of amplicons has been developed, that may provide in specific cases advantages in cost efficiency and flexibility over current qPCR approaches.for the improvement detection and identification an optimised strategy was developed to identify unknown GMOs starting from unexplained GMO elements in a given sample. The new approach yielded sequence information that is likely to be related to the model-UGMO, indicating the proof-of-principle of this method. The data were analysed based on a newly developed (WP1) NGS data analysis pipeline.

With relation to customs issues, for the identification of tobacco in reconstituted tobacco samples, LAMP- as well as real-time PCR methods have been developed and validated. With relation to the profiling of Tobacco batches, it was concluded that RAD sequencing could be used for tobacco species identification, possibly including mixtures, but that the the depth of tag used in these experiments will need further elaboration. For the identification of species, including CITES, species in complex samples, that may be degraded, an informative protocol was developed that can basically identify all species present in a sample, provided that the related sequence information is present in the related databases. In addition to this a bioinformatics pipeline has been developed and made available for (customs) laboratories via a user-friendly webtool. The method has been tested in an interlaboratory validation trial.
Project Context and Objectives:
There is increasing demand for a more detailed identification of the composition of agricultural commodities and derived food and feed products. In this respect, DNA-based methods for detection and identification of products derived from biological species have gained interest due to the robustness of the DNA molecule, the vast experience with established, reliable and efficient DNA-based methods such as polymerase chain reaction (PCR) as well as, the availability of genomic sequence data from a rapidly growing number of species. In the last few decades, we have seen important developments in the DNA- based methods. Methods were developed for identification of individual species and varieties of animal, vegetal or microbial origin and also for more specific purposes. Examples are the identification of genetically modified organisms (GMOs), molecular- epidemiological typing of pathogens implicated in disease outbreaks and wildlife forensics by customs laboratories curtailing illicit trade in endangered species. At the same time, important progress has been made in terms of:
• Setting quality criteria for new DNA-based methods, formulated as minimal performance parameters;
• Standardisation of methods, including validation procedures;
• The availability of reference materials or use of databases that are based on internationally recognised taxonomic standard materials, for instance the BOLD (barcoding of life) database;
• Extension and improvement of the biochemical toolbox (polymerases, ligases, etc.).

At the same time it is clear that there is still much need for developmental work in this field, not in the least to reduce the costs that are inherently linked to the performance of many different analyse to identify fraudulent or even hazardous ingredients in our food supply chains. Recent developments indicate that DNA sequence analysis may in many instances provide more informative and indeed more cost-effective alternatives to many current strategies of DNA analysis. With increasing demands for effective (multi)methods and on-site applications, it is important that international, joint initiatives take the lead to avoid the replication of dispersed small-scale efforts and to ensure synergy, harmonisation and standardisation.

The Decathlon project has brought together a broad range of experts and expertise to jointly work on the development of new or improved methods that are needed in the field of 1) food pathogens, 2) traceability of GMOs and 3) customs issues. The project has developed advanced methods for all three application areas with method characteristics that meet the requirements of the individual areas. Field-related technical expertise has provided insight into the most relevant bottlenecks in the fields of application, that has guided the choices for methods to be developed within the project.

Project Results:
The work in WP1 has more recently focused on isothermal amplification, digital droplet PCR (ddPCR) and Next Generation Sequencing (NGS). Within WP1 experimental work is performed to aid especially WP2, 3,4 and 5 that further develop the respective techniques in the different areas of application. In WP1 work has been performed on isothermal amplification to underpin the work in WP2 (on-site detection) and WP3 (food pathogens). In WP2 the initial focus is on on-site methods for food pathogen detection and it is important to have isothermal amplification methods available. In WP1 a review has been performed of available isothermal methods for the detection and identification of food pathogens with the emphasis on detection of EHEC and the different target genes were identified. Published assays were assessed in terms of target genes, sample matrices tested, type of amplification product detection, validation data available, etc. In addition to this also ddPCR has been explored as an alternative method for PCR-based quantitation of DNA samples. For the development of multiplex ddPCR systems to quantify GMOs two different analysis strategies have been followed: (i) quantification of twelve EU authorized GM maize lines in a single analysis and (ii) screening for NOS terminator positive and 35S promotor positive samples with conventional qPCR (first step) and ddPCR based quantification of all positive samples for the remaining four EU authorized GM maize lines (second step). As this qPCR modules were designed for use in simplex qPCR reactions, the primers and probes had to be checked carefully for interactions and the ddPCR methods have been optimised accordingly. For NGS a comprehensive report has been written to give an overview of the state-of-the-art of NGS (Next Generation Sequencing) strategies as reported in the scientific literature in the areas of detection and identification of food pathogens, of GMOs and in the field of species identification. Under WP1 all bioinformatics specialists work together in the Task Force Bioinformatics, that has progressed on the development of user friendly modules and pipelines for the routine analysis of whole genome and DNA target enriched datasets for the different Decathlon end-points.

WP2, focusing on on-site applications of DNA-based methods, is divided into two main functions. The first task encompasses development of a microfluidic cartridge for spatial multiplexing of an existing DNA molecular method. This involves discretising a purified DNA sample into a series of equal volumes. Currently the cartridge is performing as planned and can split a DNA sample into the required aliquots. It is scalable; a number of variants have been tested with between 4 and 18 separate wells. Relative to the manufacturing technique used it shows a high degree of reliability. Development of the ‘spin-stand’ test bed is currently being completed. The microfluidic disc has been tested using LAMP primers (provided by WP3) using purified DNA samples (provided by WP3). Furthermore, the issue of the purification of DNA from lysed samples to a quality required for LAMP amplification has been addressed. Based on input from WP3 it was determined to use a silica bead based purification method. A liquid handling protocol was determined. Currently the disc has been designed and initial versions have been tested. Flow visualisation (using stroboscopic imaging) shows the disc is performing as designed. However, minor iterative improvements are required before testing with samples (DNA isolates and later different food matrices) is initialised. As the spin-stand is not yet complete, LAMP amplification has been attempted using a custom heating jig and off-disc fluorescent measurements. However, this phase of testing is at an early stage. With respect to the graphene-based sensor approach. Here, it was found that nanogaps, manufactured under vacuum on free standing graphene membranes, are folded or ruptured when exposed to a water environment. Novel manufacturing methods have been developed to circumvent this issue and thus permit the nanogaps to be deployed in an aqueous solution. Work has focussed here on determining the biomolecular detection limit. This initially involves electrical characterisation of the DNA sensors. An alternative fabrication protocol was designed for, in the end, supported graphene nanogaps. Also, two different protocols to transfer the graphene monolayer on top of the pre patterned structures were tested. Further functionalization of the edge of the etched structure will be studied.

In WP3, focusing on food pathogens, the focus is on the improvement of EHEC diagnostics, using NGS and DNA sequence-based approaches for diagnostic purposes and comprehensive risk assessment of food pathogens. Also, experiments were performed to develop and optimise modules for the extraction of DNA directly from larger volumes of food matrices (minced meat or vegetables) or for the isolation of intact bacteria from irrigation water, with the aim to provide DNA of sufficient quality and quantity to be successively analysed with molecular analytical modules for the detection of the pathogen / virulence factors. The meat/vegetable protocol was tested on samples spiked at concentrations challenging the theoretical limit of detection, and demonstrated significantly improved detectability with recovery rates ranging around 15-20% and 40-50% for meat and vegetables, respectively. Furthermore, two bioinformatics workflows for genome marker identification from larger sets of bacterial genome sequences were developed and evaluated. Based on the comparative genomic analyses, 47 new specific biomarkers for the clinically most relevant STEC subtypes (O157 and non-O157 ‘big six’ serotypes as well as for O104). A multiplex PCR was developed based on the newly discriminative biomarkers, which allow STEC detection and discrimination of the clinically most relevant STEC serotypes as mentioned before. The specificity of the multiplex PCR was shown based on well defined clinical STEC isolates as well as several established E. coli reference strain collections, including the ECOR, DEC and HUSEC collections. In summary, an improved DNA-based typing tool for clinically relevant STEC has been developed, that can not only be used for improved typing and risk assessment in clinical samples, but also in food-related samples. This research will thus contribute to the further improvement of detection and typing approaches of pathogens, as well as the integration of foodborne pathogen detection and subtyping into a single test.
WP4 focused on improved methods for GMO detection and identification, including unauthorised GMOs. A ddPCR protocol for the identification and quantification of all EU-authorised GM maize varieties, which was prepared in WP1, was tested in three different labs of partners. The results of the transferability study have shown that the protocol was successfully transferred to other labs. The in-house validation of the method with two multiplex reactions met all the minimum performance parameters as established in WP6. Therefore the method was selected by WP1 as the method suitable to enter into a full collaborative trial validation lead by WP6. Additional to the performance parameters, the cost-effectiveness of the multiplex ddPCR method was assessed in line with current GMO quantification procedure with qPCR. It was shown that the multiplex ddPCR for quantification of twelve maize lines was more cost effective than qPCR in all cases but one, which is the negative sample, where only screening is done with qPCR. Furthermore, NGS approaches have been initiated. A method for NGS analysis of amplicons has been developed, that may provide in specific cases advantages in cost efficiency and flexibility over current qPCR approaches. It was shown that the NGS PCR amplicon sequencing provides generally the same screening information as qPCR screening. In general it was concluded that the flexibility of NGS amplicon targeting requiring less specificity testing seems currently the biggest benefit of this strategy. Then, for the improvement detection and identification an optimised strategy was developed to identify unknown GMOs starting from unexplained GMO elements in a given sample. The new approach yielded sequence information that is likely to be related to this model-UGMO, indicating the proof-of-principle of this method. The data were analysed based on a newly developed (WP1) NGS data analysis pipeline.
Under WP5, the customs issues are: i) identification of tobacco in reconstituted tobacco, ii) tobacco profiling (comparison of blends, batches, manufacturers) and iii) identification of (endangered) species in complex samples (multiple ingredients). For the identification of tobacco in reconstituted tobacco samples, LAMP- as well as real-time PCR methods have been developed. Out of the assays tested, the PMT-1 target was selected for prevalidation experiments. Based on these experiments, a validation trial was planned and performed for the tobacco specific detection using the PMT-1 real-time PCR assay. Together with this real-time method validation trial, it was planned to collect data for a LAMP assay, as LAMP is a potential on-site method that could be used on-site by, for instance, customs inspection services. The LAMP Nia-Green assay turned out to be the most promising method to test in parallel with the real-time PCR method. Based on these results it was concluded that RAD sequencing could be used for tobacco species identification, possibly including mixtures, but that the the depth of tag used in these experiments is not good enough to yield reliable SNP markers for effective species identification in mixtures. The second issue has to be taken into consideration is that the tag size (33-36 bp) is relatively small. Therefore, the approach is generally prone to yield false positive SNPs when mapping to the scaffold. It was therefore concluded that additional strategies will need to be explored to make it feasible to identify subsamples of individual tobacco batches. For the identification of species, including CITES, species in complex samples, that may be degraded, an informative protocol was developed that can basically identify all species present in a sample, provided that the related sequence information is present in the related databases. In addition to this a bioinformatics pipeline has been developed and made available for (customs) laboratories via a user-friendly webtool. After elaborate testing it was concluded that almost all species in the tested mixtures could accurately be identified at the species or family-level using the proposed DNA metabarcoding approach, even in cases where the larger part of the ingredients (species) were present at 1% dry weight concentration.
WP6 has compiled a comprehensive set of minimum performance parameters (MPPs) for molecular analytical methods together with their associated acceptance values (AAVs), covering techniques for all steps from nucleic acid extraction to next generation (high throughput) nucleic acid sequencing. These MPPs and AAVs can be used to assess and improve the quality of analytical methods and services from developmental stages, via validation to routine application. Much emphasis has been put on making the MPPs and AAVs applicable across sectors and balancing pragmatism with focus on maintaining sufficient quality. Application of these MPPs and AAVs are also expected to reduce costs by efficient resource investments, e.g. in the development phase and when laboratories select among alternative analytical methods. The first method developed in the frame of the DECATHLON project and selected for validation was a digital droplet polymerase chain reaction (ddPCR) method for the quantification of 12 EU- authorized genetically modified (GM) event. However, after the prevalidation, the intended validation could not be completed as a consequence of a long series of unfortunate and largely unpredicted factors. Then, two methods have been developed for the distinction of cigarettes (outer wrapper of paper) and cigarillos/cigars (outer wrapper of reconstituted tobacco/ tobacco): a quantitative real- time PCR (qPCR) assay and a Loop-mediated isothermal AMPlification (LAMP) assay. Both methods have been put to the test in an international ring trial to determine their performance. Overall, it was concluded that the developed DNA analysis methods will provide useful tools for customs: the ability to detect the presence of tobacco in the outer wrapper of these products is important for the tariff-classification of cigarettes and cigarillos/cigars, with regard to import duties and excise duty rates. The development of the combined method and pipeline for CITES species detection was selected for a large inter-laboratory trial validation. The CITESspeciesDetect pipeline is developed to perform reference based parallel identification of multiple amplified barcodes and mini-barcodes from plant and animal species. The input data are paired-end (PE) sequence reads from Illumina MiSeq or HiSeq sequencers. The output is a table of hits of operational taxonomic units (OTUs) from reads mapping to reference databases. It was concluded that the pipeline is a reliable tool, that is fit for purpose. With the recent availability of the online web-interface it became possible to re-analyze the data by each of the individual participants of the inter-laboratory trial.

Potential Impact:
Please provide a description of the potential impact (including the socio-economic impact and the wider societal implications of the project so far) and the main dissemination activities and the exploitation of results. The length of this part cannot exceed 10 pages.

Potential impact in the three areas of the project: 1) food pathogens, 2) GMOs (genetically modified organisms, with a focus on unauthorised GMOs), and 3) customs issues (with a focus on tobacco and endangered species identification):
In the field of food pathogens, the focus has been on the improvement of EHEC diagnostics, using NGS and DNA sequence-based approaches for diagnostic purposes and comprehensive risk assessment of food pathogens. Optimise modules have been developed for the extraction of DNA directly from larger volumes of food matrices (minced meat or vegetables) and for the isolation of intact bacteria from irrigation water. Especially for the food applications, the results were very promising: the meat/vegetable protocol was tested on samples spiked at concentrations challenging the theoretical limit of detection, and demonstrated significantly improved detectability with recovery rates ranging around 15-20% and 40-50% for meat and vegetables, respectively. This allow a more efficient screening for pathogenic EHEC strains in food matrices.
Further to this, two bioinformatics workflows for genome marker identification from larger sets of bacterial genome sequences were developed and evaluated. These workflows can also be applied to similar questions related to other groups of bacteria. Based on the comparative genomic analyses, 47 new specific biomarkers for the clinically most relevant STEC subtypes (O157 and non-O157 ‘big six’ serotypes as well as for O104). A multiplex PCR was developed based on the newly discriminative biomarkers, which allow STEC detection and discrimination of the clinically most relevant STEC serotypes in a very efficient way. The specificity of the multiplex PCR was shown based on well defined clinical STEC isolates as well as several established E. coli reference strain collections, including the ECOR, DEC and HUSEC collections.
In this way, an improved DNA-based typing tool for clinically relevant STEC has been developed, that can not only be used for improved typing and risk assessment in clinical samples, but also in food-related samples. This research will thus contribute to the further improvement of detection and typing approaches of pathogens, as well as the integration of foodborne pathogen detection and subtyping into a single test.
In addition to this a disc device has been developed that will allow on-site detection of these markers using derived LAMP methods. In the project the proof-of-principle has been shown, i.e. the disc device has been set up in a sample-to-answer set-up. Based on the sequence information of the newly identified markers LAMP methods can be developed for point-of-need screening. The disc device is a versatile device that may be applied for many other on-site applications for which isothermal methods are available.
With relation to GMO analysis, the project focused on improved methods for GMO detection and identification, including unauthorised GMOs. A ddPCR protocol for the identification and quantification of all EU-authorised GM maize varieties, was prevalidated in three different labs of partners. The results of the transferability study have shown that the protocol was successfully transferred to other labs. The in-house validation of the method with two multiplex reactions met all the minimum performance parameters as had been established within the project.
Additional to the performance parameters, the cost-effectiveness of the multiplex ddPCR method was assessed in line with current GMO quantification procedure with qPCR. It was shown that the multiplex ddPCR for quantification of twelve maize lines was more cost effective than qPCR in all cases but one, which is the negative sample, where only screening is done with qPCR. The ddPCR strategy can in this way broaden to array of methods that are available for rapid identification of authorised GMO events in a given sample. In addition, the ddPCR allows for highly accurate quantification, also in highly processed matrices, which fulfills a clear need in current routine GMO analysis programmes.
Furthermore, NGS approaches have been initiated for further improvement of GMO analysis strategies. A method for NGS analysis of amplicons has been developed, that may provide in specific cases advantages in cost efficiency and flexibility over current qPCR approaches. It was shown that the NGS PCR amplicon sequencing provides generally the same screening information as qPCR screening when focusing on EU-authorised GMOs. Especially in those cases where unauthorised GMOs have been identified, but no reference materials are available, this highly versatile approach may have considerable added value in laboratories performing routine GMO analyses. An NGS data analysis pipeline has been developed that will allow the rapid analysis of the NGS-generated sequences. This pipeline will become available for laboratories performing routine GMO analyses.
Then, for the improvement detection and identification an optimised strategy was developed to identify unknown GMOs starting from unexplained GMO elements in a given sample. When, based on routine screening strategies, GMO elements have been identified that may be related to unidentified unauthorised GMOs, then this procedure will in most cases prove sufficient to characterise or identify the unauthorised GMO. The new approach very specifically amplifies the GMO element – related sequences, while reducing the background DNA to virtually zero, thus providing the basis for the next NGS (Next Generation Sequencing) step that will yield the relevant sequence information. In the project the proof-of-principle of this method was shown. The data were analysed based on a NGS data analysis pipeline tha was newly developed within the project and will become available for laboratories performing routine GMO analyses.
The customs issues in the project included i) identification of tobacco in reconstituted tobacco, ii) tobacco profiling (comparison of blends, batches, manufacturers) and iii) identification of (endangered) species in complex samples (multiple ingredients).
For the identification of tobacco in reconstituted tobacco samples, methods have been developed for the specific and sensitive detection and identification of tobacco-derived DNA in materials of interest. For the detection of tobacco DNA in the outer wrappers of cigarillo’s/cigars (reconstituted tobacco), LAMP- as well as real-time PCR assays have been developed and tested. PMT-1 target was chosen for pre- validation experiments, to obtain preliminary information on performance parameters and possible fitness of samples for validation experiments. Based on these experiments, a validation trial was planned and performed for the tobacco specific detection using the PMT-1 real-time PCR assay.
Together with this real-time method validation trial, it was planned to collect data for a LAMP assay, as LAMP is a potential on-site method that could be used on-site by, for instance, customs inspection services. The LAMP Nia-Green assay turned out to be the most promising method to test in parallel with the real-time PCR method.
Overall, the developed DNA analysis methods will be able to provide useful tools for customs. The ability to detect the presence of tobacco in the outer wrapper of these products is really important for the tariff-classification of cigarettes and cigarillos/cigars, with regard to import duties and excise duty rates. When the outer wrapper is made of 100% natural tobacco, the microscopic characteristics of the tobacco allows for a straight forward confirmation and additional tests are not required. However, when the outer wrapper is made of reconstituted tobacco, it is almost impossible to detect any tobacco material microscopically. In such cases an extra (eg. DNA) confirmation method is welcomed to detect the tobacco in the outer wrapper.
Optimisation and successful re-validation of the LAMP Nia-Green method is necessary in order to be fit for purpose. The MPP for the qPCR method are very close to their AAV, and the ring trial was able to pinpoint several aspects for further development. The data in this report suggest that the values for false positive and false negative rate may improve with application of a higher dilution factor. As a result of that the sensitivity and specificity will improve as well. Alternatively, the samples could be diluted to a certain (maximum) concentration, for instance 100 genome copies, corresponding to an absolute concentration of 0.44ng/reaction. With these proposed adjustments, that may require further (in-house) validation, the method is ready for routine application.
The basic idea behind the identification of blends or batches of tobacco is to develop a technique that is able to make a so called “DNA profile” of tobacco blends, that may be reproducible in several subsamples of the same batch. By filling a database with profiles of varieties and blends, reference data can be created, but also in individual cases it may be informative to be able to test whether two tobacco portion may likely be derived from the same original batch, or whether it is highly unlikely that this is the case. The hypothesis is, that by comparison of profiles of seizures, matches will occur. Three methods have been evaluated for DNA profiling of tobacco. RAD sequencing, Genome Survey Sequencing (GSS) and target enrichment Next Generation Sequencing (NGS). In all strategies, the chloroplast DNA has been the target to screen for variety specific polymorphisms, like Single Nucleotide Polymorphisms (SNPs), insertions, deletions etc.
Based on these results it was concluded that RAD sequencing could be used for tobacco species identification, possibly including mixtures, but that the the depth of tag used in these experiments is not good enough to yield reliable SNP markers for effective species identification in mixtures. The second issue has to be taken into consideration is that the tag size (33-36 bp) is relatively small. Therefore, the approach is generally prone to yield false positive SNPs when mapping to the scaffold. It was therefore concluded that additional strategies will need to be explored to make it feasible to identify subsamples of individual tobacco batches.
Then, for the identification of species, to allow the Customs laboratories to effectively perform the analysis of samples for the presence of endangered (CITES-listed) species, it was necessary to develop methods that can broadly identify any species present in a sample under investigation. Also, samples of interest may often contain highly processed and/or degraded DNA, leading to an additional requirement that developed methods needs to identify species on the basis of short DNA sequences that may still be present in the highly processed materials. To this end it was necessary to further develop concepts and methods that relate to the selective amplification of available DNA barcoding and minibarcoding markers to simultaneously identify multiple species in a given complex sample.
Based on a detailed literature review, candidate universal DNA barcode and mini-barcode markers for animals and plants were identified. The advantage of the combined use of multiple DNA barcodes in sample analysis is that it will allow for improved resolution when some markers fail to resolve. Furthermore, it allows for better quality assurance when species can be identified with more than one DNA barcode.
The markers in the DNA barcode panel were modified to have an additional Illumina tail sequence at 5’ end of the primers. This addition allows for the Index PCR protocol for NGS library preparation, which is more cost efficient compared to ligation based NGS library preparation. The DNA barcode panel was evaluated on 34 reference materials comprising of a taxonomic wide range of plant and animal species. For each reference species, at least one DNA target could be amplified using the DNA barcode panel.
For the large amount of data that are generated during NGS analysis a dedicated NGS bioinformatics pipeline has been developed and made available via a webtool. A parameter scan was performed in order to assess the effect of software settings on the ability to identify species. The evaluation allowed identifying important parameters and their effect on the sensitivity, specificity and robustness of the procedure, which resulted in specifying recommended (default) parameters values for analysing DNA metabarcoding datasets using the CITESspeciesDetect pipeline.
It was concluded that 9 out of 10 species in the tested mixture could accurately be identified at the species or family-level using the proposed DNA metabarcoding approach, 7 of which were present at 1% dry weight concentration. Although the sensitivity and reproducibility of the procedure can be improved by applying a OTU abundance threshold of zero, this will lead to false positive identifications in these and other samples, and is therefore not recommended. Based on these results it was decided that this method was accepted for a full interlaboratory validation trial.
The approach was found to be highly reproducible across laboratories, and sensitive enough to detect species present at 1% dry weight content in experimental samples containing up to 11 species ingredients. Though, not all laboratories could identify all taxa in all cases. All six animal taxa from phylogenetically unrelated orders could be identified at the species level, including the CITES listed species Crocodylus niloticus and Huso dauricus, supporting that the method may universally target a diverse range of animal species. The web-interface, was launched on September 19, 2016 and includes a “Help” function with guidance to the users. The 16 laboratories that participated in the inter-laboratory trial were asked to evaluate the CITESspeciesDetect web- interface by re-analyzing one or more datasets generated in the inter-laboratory trial. After a one month evaluation period, feedback on the user-experience was obtained via a questionnaire and webinar. A total of 9 laboratories have tested the web-interface., and these found the CITESspeciesDetect web-interface to be well-done and fit-for-purpose.
In this way, for the identification of species, including CITES, species in complex samples, that may be degraded, an informative protocol was developed that can basically identify all species present in a sample, provided that the related sequence information is present in the related databases. In addition to this a bioinformatics pipeline has been developed and made available for (customs) laboratories via a user-friendly webtool. After elaborate testing it was concluded that almost all species in the tested mixtures could accurately be identified at the species or family-level using the proposed DNA metabarcoding approach, even in cases where the larger part of the ingredients (species) were present at 1% dry weight concentration.
Furthermore, it was concluded that the pipeline is a reliable tool, that is fit for purpose and available for customs laboratories: with the recent availability of the online web-interface it became possible to re-analyze the data by each of the individual participants of the inter-laboratory trial. This could further reinforce the perception of CITESspeciesDetect as a user- friendly and reliable pipeline (demonstrated by coherence of the results with those obtained by the developing laboratory).
These developments will allow a considerable gain in efficiency in the monitoring of samples for the presence of CITES-listed species by customs laboratories.
Other factors that will contribute to the impact of the project are:



A comprehensive set of minimum performance parameters (MPPs) for molecular analytical methods hasa been compiled together with their associated acceptance values (AAVs), covering techniques for all steps from nucleic acid extraction to next generation (high throughput) nucleic acid sequencing. These MPPs and AAVs can be used to assess and improve the quality of analytical methods and services from developmental stages, via validation to routine application. Much emphasis has been put on making the MPPs and AAVs applicable across sectors and balancing pragmatism with focus on maintaining sufficient quality. Application of these MPPs and AAVs are also expected to reduce costs by efficient resource investments, e.g. in the development phase and when laboratories select among alternative analytical methods.

A series of training activities has been performed until the end of the project, including three training webinars (to assist especially customs laboratories in the performance of the analyses as part of the validation trials of the tobacco-specific qPCR and LAMP methods, as well as for the performance of the Standard Operating Procedure for the species identification, as well as to assist the laboratories in performing the NGS data analysis using the CITESSpeciesDetect webtool), one workshop on on-site methods (at the University of Leiden, as an extra day for a smaller group of trainees, after the one-day seminar on on-site methods), three exhange visits between partners by Ph.D. students (exchange NIB, Slovenia to RIKILT Wageningen University & Research, the Netherlands, exchange RIKILT Wageningen University & Research, the Netherlands to Sjanghai Jiao Tong University, China, exchange University of Münster, Germany to NIB, Slovenia), two tutorials have been developed, one for the Decision Support Tool SIGMO and one for the CITESSpeciesDetect webtool, and one workshop for, primarily, plant breeding students in Harbin, China.
These training activities will lead to the dissemination of the newly developed methods as well as to the more rapid adoption of the new methods in the daily routine in European laboratories as well as in laboratories outside of Europe. With respect to the latter, it was also important that the Chinese partner participated in the project as this will make it easier to further harmonise methods as well as monitoring strategies on a global scale.

List of Websites:
www.decathlon-project.eu