Coffee with capsicum? How data is changing gastronomy

Grigorii Bessararov, a 31-year-old chef-mixologist at Dubai’s Frank Meat & Taps, is driven by the motto “Don’t forget to try something new!” This curiosity led him to the FlavorDB database in 2023, which maps food ingredients by the flavour molecules they share. FlavorDB has allowed him to serve flavour combinations he had never considered before, such as capsicum and chocolate, which share flavour molecules and pair well together.

FlavorDB is just one example of computational gastronomy, an emerging science that combines molecular biology, data science, and artificial intelligence. This field has applications for various stakeholders, including mixologists like Bessararov, health-conscious consumers, and those seeking personalized meal plans. Some of the most intriguing flavour pairings discovered through FlavorDB include lemon and white chocolate, pineapple and snow peas, and Grigorii’s personal favorite – bourbon, bacon, and banana.

Using FlavorDB requires both an understanding of the database's functionality and a chef's imagination to transform its suggestions into appealing dishes. Associate Professor Andrea Vitaletti of Sapienza University in Rome emphasizes that computational gastronomy aims to generate insights that help experts reflect on their intuitions and judgments about food.

Dr Ganesh Bagler, the creator of FlavorDB, has compiled a database containing 936 food ingredients, each with its own set of nearly 26,000 flavour molecules. Users of the application can discover which ingredients pair well with specific ingredients by clicking on them and revealing their molecular overlaps. For instance, coffee, which contains 269 flavour molecules, shares 78 with capsicum, which contains 218. This shared molecular structure suggests that coffee and capsicum can be successfully combined in a recipe.

Bessararov’s experiments with FlavorDB have even inspired him to create innovative cocktails, such as biscuit-infused vanilla vodka, which he believes will appeal to guests. This shift in the guest's perception of what goes well together is what Grigorii finds most exciting about computational gastronomy.