AI-powered foodome characterization

Expected Outcome: Project results are expected to contribute to all the following expected outcomes: an Artificial Intelligence (AI) tool able to accelerate the analysis of complex foodomics data to speed up the identification, structural characterisation, and quantification of chemicals present in food; improving capacity and synergies in foodomics research between academia, research infrastructures and industry. Scope: Through food and beverages, the population is exposed to an immense variety of chemicals (as contaminants and as ingredients). The “foodome” is the complete set of compounds present in a food sample at a given time; it includes both endogenous biomolecules (produced through a species' metabolic processes) and exogenous compounds (natural and synthetic substances from production, processing, handling, and packaging). The current knowledge is largely confined to a narrow set of approximately 150 food components catalogued by traditional databases. These components represent only a very small fraction of the total chemical complexity present in foods, while thousands (>100,000) of compounds and contaminants remain largely unquantified and underexplored. Understanding this “nutrient dark matter” would require identifying, cataloguing and quantifying the full spectrum of substances present in our food and present in accessible, AI-ready databases. The need for such information is today a bottleneck to make the nutrition field predictive and data driven, and has…