Development Of Reduction Algorithm For Metabolic Network Model And Guidance For Pathway Design

Constraint-based genome-scale metabolic network models have played a very important role in guiding the design of metabolic pathways and predicting cell phenotypes.With the continuous growth of genomic data and biochemical reaction data,the size and complexity of genome-scale metabolic network models has significantly increased and make the analysis of the model more difficult.Therefore,it is very important to reduce the metabolic network and obtain a "lossless" metabolic network that is convenient for calculation and does not lose metabolic information.This study develops a model reduced algorithm that automatic and does not destroy the structure of the metabolic network.The reduced model is applicable to all current model calculation methods.It not only saves a lot of calculation time,but also provides researchers with reduced models that are easy to understand.Firstly,we determined reduced rules that the number of carbon and nitrogen molecules is used to represent the metabolite(such as C1N2 or C2N5),replaced the metabolite of the reaction equation with a reduced metabolite and merged the reaction equation.According to this rule,the model reduction algorithm slimGEM was developed.At the same time,another algorithm was developed to restore the calculation results of the reduced model to multiple feasible solutions of the target metabolite synthesis pathways.Nextly,we applied slimGEM algorithm to the iML1515 genome-scale metabolic network model of Escherichia coli(2712 reactions,1877 metabolites)to obtain a reduced model ECslim656(656 reduced reactions,556 reduced metabolites).We analyzed the reduced network.The metabolite combinations with the number of carbons less than 6 and lack of direct synthetic and decomposition reactions were screened out.After calculation and restoration,two feasible cycling pathways were obtained,namely,acetaldehyde 2-3-4 synthesis cycle and acetyl-coenzyme A 4-6-5 decomposition cycle.It was found that there were two absent metabolite conversion reactions(C1+C1=>C2 and C2=>C1+C1)in the Escherichia coli metabolic network that could be achieved through the circulation pathway.The reduced reaction could correspond to one or more original reactions in iML1515.In the calculation,we obtained multiple feasible solutions of the target metabolite synthesis pathway by restoring the reduced pathway calculated by the ECslim656 model.Finally,we built a compound metabolic network model(18,600 reactions and 20,744 metabolites)based on the MetaCyc database,and reduced this model using slimGEM to obtain the MCslim1422 model(1422 reduced reactions,6552 reduced metabolites).We used this model to calculate the carbon loss-free carbon fixation circulation pathway and to filter the calculated results to obtain a feasible reduced cycle pathway,c2+c1/c3+c1(1)circulation pathway.After restoration,a new feasible carbon fixation circulation pathway,oxalate 2-3-4 synthesis cycle pathway,was found.With the help of the reduction model,we can analyze the metabolic pathways and calculate circulation pathways from a new perspective,which provides biologists with new research ideas and tools,and provides rational guidance of metabolic engineering modification.