As the number of people affected by COVID-19 disease caused by the novel coronavirus SARS-CoV-2 ebbs and flows in different national and sub-national regions across the world, it is evident that our lifestyle and socio-economic trajectories will have to be adapted and adjusted to the changing scenarios. Novel forecasting tools and frameworks provide an arguable advantage to facilitate this adapting and adjusting process, by promoting efficient resource management at individual and institutional levels. Based on deterministic compartment models we propose an empirical top-down modeling approach to provide epidemic forecasts and risk calculations for (local) outbreaks. We use neural networks to develop leading indicators based on available data for different regions. These indicators are not only used to assess the risk of a (new) outbreak or to determine the effectiveness of a measure at an early stage, but also in parametric models to determine an effective forecast, along with the associated uncertainty. Based on initial results, we show the performance of such an approach and its robustness against inherent disturbances in epidemiological surveillance data. We foresee such a statistical framework to drive web-based automatic platforms to democratize the dissemination of prognosis results.All rights reserved. No reuse allowed without permission.

The continued search for intermediate hosts and potential reservoirs for SARS-CoV2 makes it clear that animal surveillance is critical in outbreak response and prevention. Real-time RT-PCR assays for SARS-CoV2 detection can easily be adapted to different host species. U.S. veterinary diagnostic laboratories have used the CDC assays or other national reference laboratory methods to test animal samples. However, these methods have only been evaluated using internal validation protocols. To help the laboratories evaluate their SARS-CoV2 test methods, an interlaboratory comparison (ILC) was performed in collaboration with multiple organizations. Forty-four sets of 19 blind-coded RNA samples in Tris-EDTA (TE) buffer or PrimeStore transport medium were shipped to 42 laboratories. Results were analyzed according to the principles of the International Organization for Standardization (ISO) 16140-2:2016 standard. Qualitative assessment of PrimeStore samples revealed that, in approximately two-thirds of the laboratories, the limit of detection with a probability of 0.95 (LOD95) for detecting the RNA was ≤20 copies per PCR reaction, close to the theoretical LOD of 3 copies per reaction. This level of sensitivity is not expected in clinical samples because of additional factors, such as sample collection, transport, and extraction of RNA from the clinical matrix. Quantitative assessment of Ct values indicated that reproducibility standard deviations for testing the RNA with assays reported as N1 were slightly lower than those for N2, and they were higher for the RNA in PrimeStore medium than those in TE buffer. Analyst experience and the use of either a singleplex or multiplex PCR also affected the quantitative ILC test results.

The COVID-19 pandemic presents a continued public health challenge. Veterinary diagnostic laboratories in the United States use RT-rtPCR for animal testing, and many laboratories are certified for testing human samples; hence, ensuring that laboratories have sensitive and specific SARS-CoV2 testing methods is a critical component of the pandemic response. In 2020, the FDA Veterinary Laboratory Investigation and Response Network (Vet-LIRN) led an interlaboratory comparison (ILC1) to help laboratories evaluate their existing RT-rtPCR methods for detecting SARS-CoV2. All participating laboratories were able to detect the viral RNA spiked in buffer and PrimeStore molecular transport medium (MTM). With ILC2, Vet-LIRN extended ILC1 by evaluating analytical sensitivity and specificity of the methods used by participating laboratories to detect 3 SARS-CoV2 variants (B.1; B.1.1.7 [Alpha]; B.1.351 [Beta]) at various copy levels. We analyzed 57 sets of results from 45 laboratories qualitatively and quantitatively according to the principles of ISO 16140-2:2016. More than 95% of analysts detected the SARS-CoV2 RNA in MTM at ≥500 copies for all 3 variants. In addition, for nucleocapsid markers N1 and N2, 81% and 92% of the analysts detected ≤20 copies in the assays, respectively. The analytical specificity of the evaluated methods was >99%. Participating laboratories were able to assess their current method performance, identify possible limitations, and recognize method strengths as part of a continuous learning environment to support the critical need for the reliable diagnosis of COVID-19 in potentially infected animals and humans.

Meat species authentication in food is most commonly based on the detection of genetic variations. Official food control laboratories frequently apply single and multiplex real-time polymerase chain reaction (PCR) assays and/or DNA arrays. However, in the near future, DNA metabarcoding, the generation of PCR products for DNA barcodes, followed by massively parallel sequencing by next generation sequencing (NGS) technologies, could be an attractive alternative. DNA metabarcoding is superior to well-established methodologies since it allows simultaneous identification of a wide variety of species not only in individual foodstuffs but even in complex mixtures. We have recently published a DNA metabarcoding assay for the identification and differentiation of 15 mammalian species and six poultry species. With the aim to harmonize analytical methods for food authentication across EU Member States, the DNA metabarcoding assay has been tested in an interlaboratory ring trial including 15 laboratories. Each laboratory analyzed 16 anonymously labelled samples (eight samples, two subsamples each), comprising six DNA extract mixtures, one DNA extract from a model sausage, and one DNA extract from maize (negative control). Evaluation of data on repeatability, reproducibility, robustness, and measurement uncertainty indicated that the DNA metabarcoding method is applicable for meat species authentication in routine analysis.

The coronavirus disease 2019 (COVID-19) pandemic presents a continued public health challenge across the world. Veterinary diagnostic laboratories in the U.S. use real-time reverse transcriptase PCR (RT-PCR) for animal testing, and many are certified for testing human samples, so ensuring laboratories have sensitive and specific SARS-CoV-2 testing methods is a critical component of the pandemic response. In 2020, the FDA Veterinary Laboratory Investigation and Response Network (Vet-LIRN) led the first round of an Inter-Laboratory Comparison (ILC) Exercise to help laboratories evaluate their existing real-time RT-PCR methods for detecting SARS-CoV-2. The ILC1 results indicated that all participating laboratories were able to detect the viral RNA spiked in buffer and PrimeStore molecular transport medium (MTM). The current ILC (ILC2) aimed to extend ILC1 by evaluating analytical sensitivity and specificity of the methods used by participating laboratories to detect three SARS-CoV-2 variants (B.1, B.1.1.7 (Alpha) and B.1.351 (Beta)). ILC2 samples were prepared with RNA at levels between 10 to 10,000 copies per 50 microliters MTM. Fifty-seven sets of results from 45 laboratories were qualitatively and quantitatively analyzed according to the principles of ISO 16140-2:2016. The results showed that over 95% of analysts detected the SARS-CoV-2 RNA in MTM at 500 copies or higher for all three variants. In addition, 81% and 92% of the analysts achieved a Level of Detection (LOD95eff. vol.) below 20 copies in the assays with nucleocapsid markers N1 and N2, respectively. The analytical specificity of the evaluated methods was over 99%. The study allowed participating laboratories to assess their current method performance, identify possible limitations, and recognize method strengths as part of a continuous learning environment to support the critical need for reliable diagnosis of COVID-19 in potentially infected animals and humans.

Food fraud, even when not in the news, is ubiquitous and demands the development of innovative strategies to combat it. A new non-targeted method (NTM) for distinguishing spelt and wheat is described, which aids in food fraud detection and authenticity testing. A highly resolved fingerprint in the form of spectra is obtained for several cultivars of spelt and wheat using liquid chromatography coupled high-resolution mass spectrometry (LC-HRMS). Convolutional neural network (CNN) models are built using a nested cross validation (NCV) approach by appropriately training them using a calibration set comprising duplicate measurements of eleven cultivars of wheat and spelt, each. The results reveal that the CNNs automatically learn patterns and representations to best discriminate tested samples into spelt or wheat. This is further investigated using an external validation set comprising artificially mixed spectra, samples for processed goods (spelt bread and flour), eleven untypical spelt, and six old wheat cultivars. These cultivars were not part of model building. We introduce a metric called the D score to quantitatively evaluate and compare the classification decisions. Our results demonstrate that NTMs based on NCV and CNNs trained using appropriately chosen spectral data can be reliable enough to be used on a wider range of cultivars and their mixes.

Detection of food fraud and geographical traceability of ingredients is a continually sought goal for government institutions, producers, and consumers. Herein we explore the use of non-target high-resolution mass spectrometry approaches and demonstrate its utility through a particularly challenging case study -to distinguish wheat and spelt cultivars. By employing a data-independent acquisition (DIA) approach for sample measurement, the spectra are of considerable size and complexity. We utilize artificial intelligence (AI) algorithms (artificial neural networks) to evaluate the extensive proteomic footprint of several wheat and spelt cultivars. The AI model thus obtained is used to classify newer varieties of spelt, processed flour, and bread samples. Additionally, we discuss the validation of such a method coupling DIA and AI approaches. The novel framework for method validation enables calculation of precision parameters for facile comparison of the discriminatory power of the method and in the development of a reliable decision rule.

Non-targeted approaches (NTAs) are being increasingly developed and adopted to detect food fraud and identify the authenticity of the food. Method validation is a critical step before bringing the NTAs to the routine, to be assured of method performance and trust that its outcomes are reliable. However, the paucity of well structured and harmonized validation strategies has been one of the hurdles, withholding the exploitation of NTAs. This report aims to describe a validation framework for methods that involve binary classification, which are prevalent in non-targeted workflows. We foresee this work to contribute to filling the current gap in the provisions for NTA method validation; perhaps also push the dialogue further to collectively resolve existing challenges.