At present time it is possible to mark out three main approaches to biochemical pathways modeling: structural, stoichiometric and kinetic. Each of this approaches is developed for specific task and has several distinctions.

The first approach is the most popular and represents structural model development describing system topology. In other words, in such model only interaction between components is taking into consideration. Physical nature of these interactions as well as stoichiometry and reaction reversibility is not being taken into account. Consequently, such models are not be able to describe experimental data and serve exclusively to account for information about system structure and topology. It essentially constricts the region of practical application of structural models, limiting it by metabolic and genetic pathway reconstruction.

Stoichiometric modeling may be considered as more fit for description of biochemical systems behavior as such important system properties as stoichiometry and reaction reversibility are examined here. Methods Flux Balance Analysis and Elementary Flux Modes popular among researchers may be considered as examples of this approach. However these methods are far from perfect, as they describe only stationary fluxes distribution in biochemical system and do not take into account regulation realized in it. If say about applied tasks circle being solved by means of such approach then it go to the one problem - finding the gene which knockout is necessary for maximization of one or another flux. Undoubtedly, it is often the most important task for metabolism researchers, and then stoichiometric model significantly promote its solution. However, in the most cases the problem turn out much more difficult and more power instrument is needed for its understanding.

Third from selected modeling types - kinetic, may be considered as most detailed approach for several reasons at once. Firstly, during kinetic model development the researcher use all modeling methods, i.e. during reconstruction of pathway into consideration develop structure model, and during metabolic matrix composition – stoichiometric. Secondly, such models describe dynamics of system. Thirdly, all kinds of regulation, realizing in the system taking into account in them. Fourthly, by means of kinetic models it is possible to describe practically all types of qualitative experimental data, received during research of metabolic and genetic systems (metabolomics, transcriptomics, proteomics data in vitro about system components).

All these facts make kinetic modeling the most attractive during biochemical systems research. However there are some difficulties here. Thus, for two first types of modeling not only special knowledge is required but also ability to structure biological information and use corresponding software. Whereas for kinetic models with high prediction power development is required long and laborious work taking much more time then subsequent analysis of its behavior and predictions.

• Strategy of kinetic modeling of large-scale biochemical systems