Knowledge gaps in food composition are driven by the diverse nature of food itself, including variations caused by how it is produced, stored and distributed, and processed, coupled with technological and accessibility barriers1. Within the planet’s edible biodiversity, an estimated 26,000 biomolecules occur, the overwhelming majority of which are unidentified and whose health effects are generally unknown, representing the ‘dark matter’ of nutrition2.

Several existing databases strive to compile wider information about the nutritional composition of food biodiversity. The International Network of Food Data Systems (INFOODS) food composition database for biodiversity, for example, goes beyond the cultivated species level to emphasize variety, cultivar and race variation, as well as wild foods3. The US Department of Agriculture (USDA) FoodData Central’s foundation foods4 has integrated metadata on genetics and environmental factors including management and processing for foods in the USA. WikiFCD enables users to pursue research questions and projects that are currently difficult to explore5. These important advances noted that existing food composition resources generally focus on analysing major dietary staples with data on only one or a few cultivars and typically highlight no more than 150 biomolecules and key macro- and micro-nutrients.