nai

Project Progress

Part 1: Compartmentalization and interactions in ecosystem-level metabolism
In order to help understand the evolution and dynamics of metabolism in microbial ecosystems we developed an extension of genome-scale flux balance models of metabolism to multiple interacting species. Since the detailed distribution of metabolic functions among ecosystem members is often unknown, it is important to investigate how compartmentalization of metabolites and reactions affects flux balance predictions. As a first step in this direction, we addressed the importance of compartmentalization in the well characterized metabolic model of the yeast Saccharomyces cerevisiae, which we treated as an “ecosystem of organelles”. In addition to addressing the impact that the removal of compartmentalization has on model predictions, we showed that by systematically constraining some individual fluxes in a de-compartmentalized version of the model we can significantly reduce the flux prediction errors induced by the removal of compartments. We expect that our analysis will help predict and understand metabolic functions in complex microbial communities and provide novel insight on the evolution of endosymbiosis and multicellularity.

Part 2: Arithmetic simplicity beneath metabolic network architecture
Is the evolution of metabolism subject to general principles, beyond the unpredictable accumulation of multiple historical accidents? We searched for signatures of such optimality in an idealized artificial chemistry model, where it is feasible to systematically explore a complete set of efficient metabolic pathways of minimal length between any two compounds. These pathways display a modular organization of recurring topologies, including autocatalytic cycles, and a logarithmic dependence of pathway length on input/output molecule size. Across all pathways, we predict the emergence of ubiquitous metabolites, and a broad spectrum of reaction utilization, with certain reactions serving as universal steps. Similar properties hold for real metabolic networks, suggesting that optimality principles and arithmetic simplicity underlie biochemical complexity.