From Farm to Water to Patient: Genomic Source Attribution of Shiga toxin-producing E. coli.

Human infections with Shiga toxin-producing Escherichia coli (STEC) are rising in the Netherlands and are classified among the zoonoses of highest current risk. Despite decades of investment in food safety surveillance, only about 40% of infections can be linked back to food, leaving the majority of infections without an identified source. This gap points to overlooked environmental pathways, particularly livestock reservoirs and contaminated surface waters, that remain poorly understood. My project will generate the first national, year-long dataset of STEC genomes from humans, underrepresented livestock, and surface waters, combining these data with case-control epidemiology to quantify the contribution of environmental exposures. By applying supervised machine-learning approaches to genomic data, I will be able to attribute infections to their sources with greater accuracy and identify risk factors tied to specific reservoirs rather than broad ecological associations. This integrated approach will move beyond piecemeal surveillance and instead provide a unified picture of how human, animal, and environmental compartments interact to sustain transmission. The results will provide evidence to inform water-quality standards, manure management, and farm biosecurity, while building a model of genomic attribution that can be integrated into national surveillance. In doing so, the fellowship aligns directly with European priorities on zoonotic disease preparedness and sustainable agriculture, ensuring that the outcomes will be relevant not only for the Netherlands but also for other countries facing similar risks. Through this work I aim to close the largest blind spot in STEC epidemiology in the Netherlands, producing insights that are directly relevant for public health policy, food safety, and One Health preparedness across Europe.