Enzyme cost minimization


Main  |  Model  |  Workflow  |  Data and software  |  Project "Enzyme dynamics and function"



Escherichia coli model As an example, we constructed a kinetic metabolic model of E. coli central metabolism and predict its metabolite and enzyme levels. The enzyme profiles are shaped by factors such as catalytic constants, thermodynamic driving forces, enzyme saturation, and allosteric regulation, and being energetically unfeasible turns out to be a limiting case of high enzyme cost. Balance relations between the costs of adjacent enzymes provide a link to metabolic control analysis. Aside from predicting the enzyme and flux costs, ECM also leads to general flux cost functions, which can be used to bridge the gap between constraint-based and kinetic metabolic models.

Enzyme Cost Minimization was used to build and analyze a model E. coli central metabolism. Below, the results can be found in the SBtab table format (for models and data) and SBML (for models only).

Data We ran ECM three times based on different assumptions about protein burden: uniform cost (same cost for each enzyme molecule); size-dependent cost (cost proportional to chain length); and composition-dependent cost (accounting for the different ATP investment in different amino acids). It turned out that these choices have little effect on the prediction accuracy.

The NEOS files can be used for running optimisations on the NEOS Optimization Server: please unzip the file and upload the individual files to the NEOS website for enzyme cost minimization. The fields for files "moietymet", "moietyval", "alpha", and "A" can be left open.

Uniform protein cost
Results [.pdf]
Input data
Network model [SBML]
Model and data [SBtab]
Model and data [NEOS files]
Result files
Model and state (EMC4cm) [SBtab]
Metabolic state (EMC..) [SBtab]
Size-dependent cost
Results [.pdf]
Input data
Network model [SBML]
Model and data [SBtab]
Model and data [NEOS files]
Result files
Model and state (EMC4cm) [SBtab]
Metabolic state (EMC..) [SBtab]
Composition-dependent
Results [.pdf]
Input data
Network model [SBML]
Model and data [SBtab]
Model and data [NEOS files]
Result files
Model and state (EMC4cm) [SBtab]
Metabolic state (EMC..) [SBtab]