Metabolic Network Side Path Calculation Method For The Research

Cellular metabolism is the basis of all life activities. The interaction between the reactions of metabolic network maintains the cell’s metabolism. Biological metabolic network is a dynamic, orderly network. The imbalance of the metabolic network will lead to the occurrence of many diseases. Now, with the advances in high-throughput experimental techniques and the new sequencing technology, mathematical model based on the COBRA framework for genome-wide metabolic networks are constantly being improved. The mathematical analysis of complex metabolic networks is taking more and more important role. In mathematics, extreme pathway of metabolic network model are described as the edge of the solution space, these extreme pathways can non-negatively linearly represent any point of the solution space. Extreme pathways are independent and minimum metabolic pathway in the metabolic network in the biological sense. Any metabolic pathways in the entire metabolic network can be expressed by them. Extreme pathway analysis can be applied to metabolic engineering, drug targets studies, etc. Extreme pathway analysis can make the metabolic network analysis more targeted and more efficient. Of course, calculation of the extreme pathway is the basis of the study. Therefore, the study of the method to compute extreme pathway has great significance.In this paper, two new methods to calculate extreme pathway had been proposed. One approach is based on mixed integer linear programming, this approach use the "corner point" sampling results as the initial point of mixed integer linear programming method. At the same time, the length of extreme pathway is the objective function of MILP to optimize the programming. Another method is based on the principle of simplex method. This method was calculated based on the theory of the rotation of vertices in the convex space. These two methods can overcome the deficiencies of the existing calculation method in the calculation of extreme pathway in large-scale metabolic network. The comparative analysis of these two new methods with existing methods had shown the broad applicability and flexibility of this method.The two new methods proposed in this paper can be applied to any metabolic network, especially the whole genome scale metabolic network. These methods require few parameters:the stoichiometric matrix of metabolic network modeled by COBRA framework and the reaction reversibility information. By the way, these two methods can also be used to calculate the specific extreme pathways contain the given reaction sets. These pathways have important guiding significance in the study of the reaction’s importance, redundancy of network and many other researches.