Professor Séamus Fanning
Food manufacturing and processing facilities contain millions of different bacteria, many are neither harmful to food nor to human health. However, a food quality and food safety risk is triggered when harmful bacteria, which can spoil food or pose a threat to human health, enter food production facilities. Current methods used to control such bacteria are neither sufficiently rapid nor specific.
They also use large amounts of energy, water and chemicals none of which are sustainable or kind to our environment. The SAFE programme aims to develop a new state-of-the-art food safety and quality decision making software toolbox to mitigate against the risk of bacterial contamination in the food supply chain in a smarter, faster and in a more specific and sustainable way.
During a 2-year period researchers at UCD tracked the environments in a number of food manufacturing plants in Ireland belonging to the industry partners. These plants include infant formula grade ingredient plants, a cooked and fermented meat processing plant and a precision vitamin and mineral pre-mix manufacturing facility. Seasonal and climate changes will be taken into consideration during this period as such changes can cause shifts in the microbial communities or “microbiome” of the facilities.
These changes affect food quality, safety and the nutritional profile of the final product. By mapping these microbiomes across the seasons the consortium will develop databases which leverage gene sequencing technology and statistical analysis to define bacterial characteristics at the DNA level. These databases will then be used to develop a predictive software toolbox. This toolbox will enable quicker and more accurate quality control analysis of the bacteria present in food facilities. This will prevent bacteria which can spoil food or pose a human health risk entering the food supply chain in a faster and a more sustainable way.
By developing a state of the art safety and quality decision making toolset to mitigate the risk of contamination in the food supply chain, this project demonstrates a new level of partnership, collaboration and joined up thinking between our client companies and our research institutes.
With a huge amount of data available from just a single sample, the approach was typically to look just at several bacteria of interest. The output from the sequencing is a count on the number of reads per bacteria (OTUs). The counts will vary based on sample quality so we focus on relative abundance rather than absolute counts. Sample can be viewed at multiple levels of the taxonomic hierarchy, from phylum down to genus. Next Generation Sequencing (NGS) techniques can facilitate the monitoring and detection of both culturable and non-culturable bacteria. This overcomes the challenge of traditional techniques which focus on a single bacteria per test. NGS generates a rich amount of data which can be leveraged through by using a machine learning pipeline to identify what the key predictors of a particular bacteria are and then train a model to predict the occurrence. The old one size-fits-all approach to microbial safety is no longer really sufficient and using the power of sequencing, big data and data science, every plant should be able to tailor its approach to address the characteristics unique to it.