Genome-Scale Metabolic Models And Flux Balance Analysis
Understanding the metabolic activities of organisms is very important to elucidate the essence of life. Various techniques have been developed to investigate the metabolic behaviours of organisms. These techniques can be used to identify the metabolic activities of organisms under different conditions. Flux Balance Analysis (FBA) is one of the powerful techniques that can be used to analyse the metabolism of organisms in silico.
Metabolism is a complex network that consists of all enzymatic reactions that take place inside a cell. Thousands of metabolites are linked to each other via reactions that convert one metabolite to another, and thousands of enzymes, which are produced based on the instructions coming from mRNAs, catalyze these reactions.
Development of systems biology techniques has made it possible to systematically analyze biological entities by computers. Systems biology approach is a systemic analysis of biological entities using biological data and computational techniques. Advancement in high-throughput biological data generation and in informatics technology has enabled systematic simulation of a cell by reconstructing Genome-Scale Metabolic Networks (GMNs). GMNs are mathematical representations of cell with the help of matrices called stoichiometry matrix, and it consists of reactions, metabolites and genes. GMNs are constructed based on gene-protein-reaction (GPR) rules that determine the link between reactions and genes through enzymes, and they enable simulation of metabolic phenotypes of various conditions by integrating transcriptome data.
Mapping transcriptome data on GMNs is widely used as an effective systems biology approach to elucidate the metabolic activities of organisms based on conditions. GPR rules can be used to link gene expression data and reactions. Using biological data and GPR rules, condition-specific GMN models can be generated by removing reactions that are not active in the condition of interest.
Constraint-based modelling (CBM) is a beneficial technique that helps to analyze complex systems of metabolism, and it was used in numerous studies so far. CBM operates under physiochemical constraints with a steady-state assumption. Mass balances, thermodynamic reliability, substrate availability are the main physiochemical constraints in CBM to simulate the metabolism of interest. The major constraint of CBM is the steady-state assumption, which assumes that the concentrations of intracellular metabolites do not change for a sufficiently long time. Although determining a relevant objective function is challenging for higher organisms as most cell types have their own unique metabolic properties, CBM studies have commonly used various objective functions that are thought to be the main goal of the metabolism of interest.
The most widely used constraint-based computational approach for the analysis of GMNs is Flux Balance Analysis (FBA). FBA is a mathematical optimization technique that uses linear optimization to predict distribution of fluxes at steady state by employing the stoichiometric matrix (S matrix). FBA uses an objective function besides the aforementioned constraints to select an optimum point from the flux solution space. ATP generation and biomass maximization are commonly used as objective functions. The major components of FBA are S matrix, objective function and additional constraints. S matrix is generated using mass balance equations around intracellular metabolites as a function of reaction rates, and Smxn consists of m metabolites and n reactions. Defining a proper objective function is crucial for meaningful results. Additional constraints play a critical role in narrowing solution space to meaningful results, and they can be determined based on wet-lab experiments. Accumulation of metabolites can be represented in mathematical form as dx/dt, where x is the vector of metabolite concentrations. The mathematical equation for the accumulation of metabolite concentrations can be written in terms of multiplication of S matrix with the vector of fluxes, which is represented as v in Equation 1. The assumption of the steady-state condition turns the complicated ordinary differential equation-based problem into a much more straightforward linear system of equations, as shown in Equation 2.
The last step of FBA is performing optimization using the necessary aforementioned constraints.
