RBA logo   Resource Balance Analysis


Main  |  What is RBA  |  Models  |  Software  |  Contact

Overview   -   Theory   -   Literature

What is Resource Balance Analysis?

Resource balance analysis (RBA) is a framework for models of living cells, assuming that cells use their internal resources economically. The approach considers a biological system at the system level and composed of interacting subsystems; it is therefore a systems biology approach at the intersection of biology, cybernetics and control theory. RBA models accurately capture the macromolecular composition of cells and can be refined step by step by adding cellular processes (such as translation, transcription, secretion, or chaperoning).

RBA model structure RBA model constraints
RBA model structure     RBA model constraints

The RBA framework, which builds whole-cell models based on metabolic network reconstructions, was developed to describe bacterial populations at steady growth and predefined external nutrient concentrations. It has since been extended to dynamical conditions, to bacterial populations in a chemostat, and to eukaryotic organisms such as plants.

But aren't RBA models complicated?

Yes, RBA models are complicated. The reason is that they do not just describe metabolism (like the popular Flux Balance Analysis models), but an entire cell, including many details such as the amino acid composition of proteins, transport processes within a cell, and the composition and efficiency of molecular machines. When building a model, all this knowledge must be gathered.

Our tools are made to simplify this process: starting from a pre-existing metabolic model (for instance, a network that would usually be used for Flux Balance Analysis), they add information (about protein sequences and typical molecular machines like the ribosome) automatically, resulting in a first consistent, but not yet fully realistic model. Further details (e.g. other molecular machines, bounds on compartment sizes) can be added manually, but these additional steps require detailed knowledge about a cell and cannot be easily simplified.

Another step in model building concerns parameter fitting: the efficiencies of enzymes and molecular machines are described by parameters called apparent catalytic constants, which need to be individually determined for each machine, and possibly as a function of cell growth. This estimation can be partially automated, but again involves some manual steps and requires knowledge about the cell to be modeled.

At the end of this model construction, the complete and parameterized model can be stored (and given to colleagues) in a special XML format called RBAml. All the following steps are less difficult!

How can I run an RBA simulation?

Running simulations with an existing RBA model is fairly easy. Given a model in RBA XML format, simulations can be run by using the RBApy or RBAtools packages. Pre-built models for different organisms can be found on our models page. For more details about simulations, see Software.

How can I build and calibrate a new RBA model?

This is the more difficult part. RBA models are built based on existing metabolic network models (the type of models used for Flux Balance Analysis). As explained above, building and calibrating an RBA model by RBApy is a stepwise procedure that is partially automated, but requires manual adjustments and takes some time. For an overview, please see the RBApy package documentation. If you have questions, please get in touch!

How does RBA relate to other modeling approaches?

Flux Balance Analysis (FBA) predicts metabolic fluxes from a metabolic network and a biomass reaction with a fixed composition. RBA starts from the same kind of network, but also accounts for the cost of producing the enzymes and molecular machines that carry these fluxes. The biomass composition is therefore not fixed in advance, but predicted, and it changes with the growth conditions. Enzyme-constrained metabolic models (such as GECKO models) limit metabolic fluxes by a total enzyme budget, but usually do not describe the production of enzymes or other cellular processes such as translation. ME models (models of metabolism and macromolecular expression) have a scope similar to RBA and usually require high-precision solvers. RBA describes cellular processes by a moderate number of molecular machines with apparent efficiencies, so that a model can be solved as a series of linear programs with standard solvers.

Learning more about RBA