The ability to obtain large amounts of heterogeneous data on the components of an animal cell using high-perfor- mance molecular biological technologies necessitates the development of computer data processing methods. One of these methods is the representation of an animal cell as a single biological system through a stoichiometric model of the metabolic network of the cell. The process of cell metabolism over time is considered in such a model as a series of sequential quasistatic states. The model validation procedure includes a number of successive stages: analysis of the metabolic network; confirmation of the conclusions of model analysis by experimental data on a living cell; tuning of model parameters aimed at more accurate imitation of cell metabolism. The model is represented by a directed graph and a mathematical matrix, which display the simultaneous state of stoichiometric equations of enzymatic reactions forming the metabolic network of the cell. When determining the functional state of the metabolic system of a cell, matrix calculations are used; the tasks of optimizing the metabolic functions of a cell are solved by linear program- ming methods and graph theory. An example of the practical application of human cell metabolism models are Recon models. They found practical application in research when determining biomarkers of the action of biologically active substances, studying birth defects in metabolism, identifying side effects of a drug action, determining targets for expo- sure to biologically active substances, and studying cancer cell metabolism. Today, metabolic models of various cells have been created for use in various fields of biomedical research: hepatocytes, cardiomyocytes, astrocytes, kidney cells, adipocytes, red blood cells, blood mononuclear cells, mesenchymal stem cells, platelets, myocytes, sperm cells, enterocytes, endothelial cells, cancer cells, brain neurons.
Read full abstract