Priit Eek

The overconsumption of sugar is one of the main reasons cardiovascular diseases and type 2 diabetes have reached pandemic proportions. While the obvious solution would be to eat less and make healthier lifestyle choices, the desire for sweetness in our food remains strong. Artificial sweeteners are widely used to lower calorie intake. Several of them, however, are suspected to pose health risks of their own, and many do not taste as well as sucrose, often leaving a lingering off-taste. The development of novel sweeteners is hindered by the scarcity of measurable data on what constitutes an ideal non-caloric sweetener. We aim to develop biochemical assays to determine these metrics, facilitating the discovery of healthy, sustainable, and good-tasting sweeteners.

In traditional food systems, additives derived from petrochemicals and animal products are widely used. While these compounds may possess desired techno-functional properties, they come with environmental, ethical, health, and sustainability issues. The goal of this project is to develop alternative protein-based food additives, such as colorants and sweeteners, that meet the needs of the food industry while addressing the concerns. Rational design, structural biology, and AI methods are utilized for protein development. In collaboration with TFTAK, a precision fermentation platform is developed to produce proteins in microorganisms. Protein samples are tested in model foods. Successful prototypes are commercialized through partnerships with the local food industry and startup accelerators. The gathered experimental data is used to model relationships between protein structure and techno-functional properties, facilitating the design of novel food proteins in the future.

This project aims to develop an AI/DFT/ML (artificial intelligence, density functional theory and machine learning)-guided workflow for designing short peptide–metal complexes as sustainable bio catalysts. The work combines AI-based peptide generation, molecular modelling, quantum-chemical and machine-learning-based catalytic screening, and experimental validation of selected candidates. By computationally prioritizing promising designs before laboratory testing, the project aims to reduce experimental trial-and-error, reagent use, waste generation, energy cost and development time in greener catalytic process development.