We use data-driven life science to understand, predict and engineer metabolism. Cells distribute their limited resources over thousands of reactions, and we study how this shapes growth, production and disease by combining genome-scale and resource allocation models with omics data, machine learning and experiments. We develop widely used open-source tools for building and analysing these models, including RAVEN for model reconstruction and GECKO for adding enzyme constraints, and we contribute to community models such as Human-GEM and yeast-GEM.
These models are freely available through Metabolic Atlas, a web portal that we help develop. It lets researchers browse, search and visualise genome-scale models of humans, model animals and microbes, and explore how genes, reactions and metabolites connect.
Our research contributes to both planetary and human health. For planetary health, we interpret omics data through enzyme-constrained models to find the bottlenecks that limit microbial cell factories. We then engineer yeasts such as Yarrowia lipolytica to produce chemicals sustainably from renewable resources. For human health, we combine human and model-animal models with patient and single-cell data to study diseases such as Alzheimer’s disease, contributing to precision medicine. Increasingly, we integrate mechanistic models with AI, including large language models, to predict enzyme kinetics and to automate model reconstruction and the interpretation of omics data.
Group members
Ricardo Bisquert Alcaraz (postdoc)
Fabian Moreno Avitia (postdoc)
Renske Verkuijlen (PhD student)
Jokūbas Daunoras (MSc student)
Minji Kim (visiting researcher)
Svitlana Myroshnychenko (project assistant)