Revision as of 20:15, 4 October 2013 by Daie07 (Talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)


When developing production processes based on the use of microorganisms can be employed genome-scale metabolic models to estimate the behavior of the organism.

Organism metabolism can be modeled by a network of metabolites and enzymes that must integrate all biochemical reactions present in the organism.

A proper metabolic reconstruction process begins with the annotated genome. This information is to be processed and be drawn all reactions that are documented for the organism in question. The set of reactions that are obtained as a result be corrected iteratively using the information available in different databases and by study of the network in order to avoid the presence of false positives (and false negatives) between reactions [1].

As a result of this process you get this network of metabolites and reactions that should represent as much detail as possible the metabolic processes that take place inside the cell.

When operating these models is to distinguish between internal flows, reactions that occur within the cell, and trade flows, which occur between the cell and its surroundings. You need to know some information about these exchange reactions, in order to establish the parameters that enable outline represent the metabolic behavior of the cell. Once validated and reconstructed metabolic model, this network can be used to carry out several studies:

  • Analysis of network connectivity.
  • Comparative studies of evolution to find patterns among organisms.
  • Improved strains: addition and deletions of genes.
  • Studies of the robustness of the network.
  • Studies regulatory parameters.
  • Analysis and metabolic flux variability.

Flux Balance Analysis, from now on FBA is a methodology used for the study of biochemical networks, including genome-scale metabolic reconstructions [2]. In this project we study cell growth versus GPP production, precursor of the products that are to be synthesized.

For the application of this algorithm, we start from the definition of the variation of the concentrations of metabolites in the cell as a function of time:

The vector Xi represents the amount of all the metabolites, the size of this vector is m (the number of metabolites). While the flow through all reactions is represented by the vector vj, of size n (number of reactions). Metabolic reactions are represented by the stoichiometric matrix S(i,j), whose size is m x n. This matrix is defined such that in a system composed of m metabolites (rows) and n Reactions (columns), the entries in each column represent the stoichiometric coefficients of all metabolites involved in the reaction for that column. A negative value indicates that the metabolite is consumed by this reaction (one of the reactants) and a positive value is produced (one of the products). The value 0 appears in all those metabolites that do not participate in the reaction in question. S(i,j) is a sparse matrix, because in most biochemical reactions take part only a few metabolites. Generally, the number of reactions (n) and metabolite (m) do not match.

To solve this indeterminate set of equations, resulting in an infinite number of solutions, it is necessary to restrict this dynamic system into a static system, for it is considered that the system is in steady state. Thus, intracellular metabolite concentrations do not change with time, and are therefore balanced. By contrast, the extracellular metabolites themselves can change their concentration versus time, enabling the assimilation of substrates and product formation.

Is justified using a steady state because the transient is faster than the cell growth so that the order of magnitude of speed is not comparable with one another. This condition is:


Stoichiometric matrix S(i,j) is the known parameter, represented the metabolic network, so that the metabolic flux vector vj is unknown.

The present system remains undetermined and take on new assumptions necessary to solve it. If some restrictions are known (biological constraints, numerical constraints) these have to be imposed. With this narrows the solution space, yet provides multiple solutions.

The last step of the algorithm applied is to use some restrictions in order to establish a unique solution within the defined solution space. This can be achieved by optimizing an objective reaction.

Figure 1 shows a conceptual diagram of the resolution process used by the FBA algorithm, which is restricting the space of solutions and finally optimize the reactions to obtain an optimum flow distribution for maximization or minimization that reaction, which is called target reaction.

Figure 1. Restriction process solutions using FBA (adapted from [2])


  1. Thiele I., and Palsson B. Ø., 2010, “A protocol for generating a high-quality genome-scale metabolic reconstruction.,” Nat. Protoc., 5(1), pp. 93–121.

  2. Orth J. D., Thiele I., and Palsson B. Ø., 2010, “What is flux balance analysis?,” Nat. Biotechnol., 28(3), pp. 245–8.