Feeding genomes to Deep Learning algorithms to detect diet-based local adaptation

Thinking about selective pressures, pathogens often come first to mind. One that is less obvious but which on the other hand exerts a constant pressure is our diet. Food resources availability and diversity, along with cultural practices and major shifts in them, is a perfect recipe for local adaptation. This proposal aims at finding signals of natural selection, indicative of adaptation, following either the shift to a diet relying on agriculture or to one based on plants. I will use Artificial Intelligence approaches to scan the genomes of populations from diverse geographic areas, hence environments, that have experienced (or not) shifts in modes of subsistence. Using an interdisciplinary framework combining genomics, anthropology and computer science, I will shed light on how our diet is shaping us, potentially leading to novel ways to approach global health issues pertaining to extant diet, to which our ancestors and their genes were not well adapted to.