Application of Whole-Genome Sequences and Machine Learning in Source Attribution of Salmonella Typhimurium.

Prevention of the emergence and spread of foodborne diseases is an important prerequisite for the improvement of public health. Source attribution models link sporadic human cases of a specific illness to food sources and animal reservoirs. With the next generation sequencing technology, it is possible to develop novel source attribution models. We investigated the potential of machine learning to predict the animal reservoir from which a bacterial strain isolated from a human salmonellosis case originated based on whole-genome sequencing. Machine learning methods recognize patterns in large and complex data sets and use this knowledge to build models. The model learns patterns associated with genetic variations in bacteria isolated from the different animal reservoirs. We selected different machine learning algorithms to predict sources of human salmonellosis cases and trained the model with Danish Salmonella Typhimurium isolates sampled from broilers (n = 34), cattle (n = 2), ducks (n = 11), layers (n = 4), and pigs (n = 159). Using cgMLST as input features, the model yielded an average accuracy of 0.783 (95% CI: 0.77-0.80) in the source prediction for the random forest and 0.933 (95% CI: 0.92-0.94) for the logit boost algorithm. Logit boost algorithm was most accurate (valid accuracy: 92%, CI: 0.8706-0.9579) and predicted the origin of 81% of the domestic sporadic human salmonellosis cases. The most important source was Danish produced pigs (53%) followed by imported pigs (16%), imported broilers (6%), imported ducks (2%), Danish produced layers (2%), Danish produced cattle and imported cattle (<1%) while 18% was not predicted. Machine learning has potential for improving source attribution modeling based on sequence data. Results of such models can inform risk managers to identify and prioritize food safety interventions.

AB_SA: Accessory genes-Based Source Attribution - tracing the source of Salmonella enterica Typhimurium environmental strains.

The partitioning of pathogenic strains isolated in environmental or human cases to their sources is challenging. The pathogens usually colonize multiple animal hosts, including livestock, which contaminate the food-production chain and the environment (e.g. soil and water), posing an additional public-health burden and major challenges in the identification of the source. Genomic data opens up new opportunities for the development of statistical models aiming to indicate the likely source of pathogen contamination. Here, we propose a computationally fast and efficient multinomial logistic regression source-attribution classifier to predict the animal source of bacterial isolates based on 'source-enriched' loci extracted from the accessory-genome profiles of a pangenomic dataset. Depending on the accuracy of the model's self-attribution step, the modeller selects the number of candidate accessory genes that best fit the model for calculating the likelihood of (source) category membership. The Accessory genes-Based Source Attribution (AB_SA) method was applied to a dataset of strains of Salmonella enterica Typhimurium and its monophasic variant (S. enterica 1,4,[5],12:i:-). The model was trained on 69 strains with known animal-source categories (i.e. poultry, ruminant and pig). The AB_SA method helped to identify 8 genes as predictors among the 2802 accessory genes. The self-attribution accuracy was 80 %. The AB_SA model was then able to classify 25 of the 29 S. enterica Typhimurium and S. enterica 1,4,[5],12:i:- isolates collected from the environment (considered to be of unknown source) into a specific category (i.e. animal source), with more than 85 % of probability. The AB_SA method herein described provides a user-friendly and valuable tool for performing source-attribution studies in only a few steps. AB_SA is written in R and freely available at https://github.com/lguillier/AB_SA.

Four European Salmonella Typhimurium datasets collected to develop WGS-based source attribution methods.

Zoonotic Salmonella causes millions of human salmonellosis infections worldwide each year. Information about the source of the bacteria guides risk managers on control and preventive strategies. Source attribution is the effort to quantify the number of sporadic human cases of a specific illness to specific sources and animal reservoirs. Source attribution methods for Salmonella have so far been based on traditional wet-lab typing methods. With the change to whole genome sequencing there is a need to develop new methods for source attribution based on sequencing data. Four European datasets collected in Denmark (DK), Germany (DE), the United Kingdom (UK) and France (FR) are presented in this descriptor. The datasets contain sequenced samples of Salmonella Typhimurium and its monophasic variants isolated from human, food, animal and the environment. The objective of the datasets was either to attribute the human salmonellosis cases to animal reservoirs or to investigate contamination of the environment by attributing the environmental isolates to different animal reservoirs.

Network Approach to Source Attribution of Salmonella enterica Serovar Typhimurium and Its Monophasic Variant.

Salmonella enterica subspecies enterica serovar Typhimurium and its monophasic variant are among the most common Salmonella serovars associated with human salmonellosis each year. Related infections are often due to consumption of contaminated meat of pig, cattle, and poultry origin. In order to evaluate novel microbial subtyping methods for source attribution, an approach based on weighted networks was applied on 141 human and 210 food and animal isolates of pigs, broilers, layers, ducks, and cattle collected in Denmark from 2013 to 2014. A whole-genome SNP calling was performed along with cgMLST and wgMLST. Based on these genomic input data, pairwise distance matrices were built and used as input for construction of a weighted network where nodes represent genomes and links to distances. Analyzing food and animal Typhimurium genomes, the coherence of source clustering ranged from 89 to 90% for animal source, from 84 to 85% for country, and from 63 to 65% for year of isolation and was equal to 82% for serotype, suggesting animal source as the first driver of clustering formation. Adding human isolate genomes to the network, a percentage between 93.6 and 97.2% clustered with the existing component and only a percentage between 2.8 and 6.4% appeared as not attributed to any animal sources. The majority of human genomes were attributed to pigs with probabilities ranging from 83.9 to 84.5%, followed by broilers, ducks, cattle, and layers in descending order. In conclusion, a weighted network approach based on pairwise SNPs, cgMLST, and wgMLST matrices showed promising results for source attribution studies.

Source Attribution of Salmonella in Macadamia Nuts to Animal and Environmental Reservoirs in Queensland, Australia.

Salmonella enterica is a common contaminant of macadamia nut kernels in the subtropical state of Queensland (QLD), Australia. We hypothesized that nonhuman sources in the plantation environment contaminate macadamia nuts. We applied a modified Hald source attribution model to attribute Salmonella serovars and phage types detected on macadamia nuts from 1998 to 2017 to specific animal and environmental sources. Potential sources were represented by Salmonella types isolated from avian, companion animal, biosolids-soil-compost, equine, porcine, poultry, reptile, ruminant, and wildlife samples by the QLD Health reference laboratory. Two attribution models were applied: model 1 merged data across 1998-2017, whereas model 2 pooled data into 5-year time intervals. Model 1 attributed 47% (credible interval, CrI: 33.6-60.8) of all Salmonella detections on macadamia nuts to biosolids-soil-compost. Wildlife and companion animals were found to be the second and third most important contamination sources, respectively. Results from model 2 showed that the importance of the different sources varied between the different time periods; for example, Salmonella contamination from biosolids-soil-compost varied from 4.4% (CrI: 0.2-11.7) in 1998-2002 to 19.3% (CrI: 4.6-39.4) in 2003-2007, and the proportion attributed to poultry varied from 4.8% (CrI: 1-11) in 2008-2012 to 24% (CrI: 11.3-40.7) in 2013-2017. Findings suggest that macadamia nuts were contaminated by direct transmission from animals with access to the plantations (e.g., wildlife and companion animals) or from indirect transmission from animal reservoirs through biosolids-soil-compost. The findings from this study can be used to guide environmental and wildlife sampling and analysis to further investigate routes of Salmonella contamination of macadamia nuts and propose control options to reduce potential risk of human salmonellosis.