Research
Our research program
We aim to connect scales, from the structure of genomes and metabolic networks, to how organisms evolve, how they interact, which communities assemble, and how they collectively function.
Is evolution of metabolism predictable?
Metabolic networks are not random wiring diagrams: thousands of reactions funnel through a small core of shared currencies and precursors. We study how this bow-tie architecture shapes which mutations matter, how their effects combine, and why fitness landscapes are so often simpler than their size suggests.
Can the chemistry of an environment tell us which community will assemble?
The nutrients available to a community are among the strongest determinants of who ends up living in it. We assemble model communities in high throughput and in bioreactors, and pair them with metabolic models, to learn how composition follows from chemistry — and how to shift it with targeted metabolic interventions.
Can we engineer what a community does without mapping every interaction?
We develop statistical approaches, heavily inspired in the theory of fitness landscapes, to predict the function of communities from their composition. We use those predictions to design consortia for diverse applications, from food fermentation to agriculture.
Papers
Structuring complexity by mapping the possible in microbial ecosystems
Current Opinion in Microbiology
Global epistasis on fitness landscapes
Phil. Trans. R. Soc. B
Can digital twins of microbial communities help us understand and engineer them?
We develop genome-scale, multi-species metabolic models to simulate microbial communities in time and space.
Papers