Nearly half of Americans have tried to lose weight in the past year, according to a CDC study, and around 80 percent have attempted dieting at some point. The high prevalence of diet-related chronic diseases such as cardiovascular disease, high blood pressure, and type 2 diabetes highlights the significant public health implications.
Despite widespread interest in nutrition, there is still much that remains unknown about food. While most consumers recognize basic nutrients like calories and some vitamins or minerals, experts estimate that a single food item can contain between 20,000 and 50,000 compounds. However, the U.S. Department of Agriculture currently tracks only about 150 chemical compounds in food. Researchers refer to this gap in knowledge as “nutritional dark matter,” which can have direct effects on health.
To address these gaps, scientists at the University of California (UC), particularly through the USDA-NIFA AI Institute for Next Generation Food Systems (AIFS) at UC Davis, are leveraging artificial intelligence (AI) to analyze data across the food system. AIFS is one of several AI institutes funded by the National Science Foundation since 2020. Its projects include FoodAtlas—a collaborative effort with UC Berkeley and UC Agriculture and Natural Resources—which uses AI to organize scientifically validated information from decades of published research.
Initiatives such as the Periodic Table of Food Initiative (PTFI) in the United States and FooDB in Canada are expanding knowledge about chemical compounds found in foods and their potential health impacts. FoodAtlas contributes by constructing a knowledge graph that connects scientific findings on food chemistry and health outcomes.
Research into traditional diets provides further insight into how specific foods may benefit health. For example, studies suggest that DMB (3,3-dimethyl-1-butanol), found in olive oil, red wine, grapes, and raisins—key components of the Mediterranean diet—may help lower risks for chronic conditions like heart disease.
FoodAtlas has already revealed that much nutritional information published in scientific journals does not make its way into accessible databases. By synthesizing data from reputable sources and linking studies with corresponding foods and compounds, FoodAtlas aims to facilitate broader understanding among researchers and clinicians.
“Knowing what chemicals are in our food and what they do to our bodies is essential for public health,” says Ilias Tagkopoulos, director of the USDA-NIFA AI Institute for Next Generation Food Systems (AIFS). “FoodAtlas has been created with this mission in mind: an AI system that reviews the published literature and databases to map foods, their molecular composition, their health effects, and other parameters that are important for decision support in creating healthier, more sustainable food.”
For consumers seeking practical guidance amid complex nutritional science, UC Davis is developing an app called “Swap it Smart.” This app will use insights from FoodAtlas and PTFI to help users formulate meals based on factors such as nutrition quality, environmental impact, affordability, flavor, bioactivity—and individual health goals. Ingredient substitutions suggested by the app will be informed by advanced research on how foods affect both human health and sustainability targets. Collaborations with California chefs are planned to validate recipes developed using this technology.
“Swap it Smart is building on the strengths of FoodAtlas and PTFI to create a platform for meal development,” Tagkopoulos says. “It is designed to help nutritionists in schools and chefs in restaurants make delicious meals and dishes that are good for our health and the environment.”
As research continues into thousands of untracked compounds present in everyday foods—including those with unknown or population-specific effects—the hope is that future dietary recommendations will become more personalized while remaining accessible.
