A method for predicting the quality of nutritional additives required to promote microbial growth based on metabolic computational model

Abstract

The present invention proposes a method for predicting the quality of nutritional additives required to promote microbial growth based on a metabolic calculation model, including obtaining or constructing a genome-scale metabolic network model of a target microorganism, which describes the metabolic network in microbial cells and calculates the flow and growth rate of metabolites. By collecting the molecular weight (MW) and kcat value of the enzyme, setting the cell protein concentration and metabolic protein ratio, introducing enzyme kinetics and protein pool data constraints, the accuracy of the model is enhanced. Using the constrained model, the exchange reaction of different nutritional additives and their uptake rate are set, the supplement specific efficiency (SSE) index is calculated, and the maximum growth rate of the microorganism after adding the nutritional additive is predicted, thereby evaluating the efficiency of the nutritional additive in improving the growth rate of the microorganism. This method is used to predict and evaluate the nutritional additives required to promote the growth of microorganisms, and significantly improves the accuracy of predicting the response of microorganisms to nutritional additives.

Description

Technical Field

[0001] The invention relates to the field of biotechnology, and in particular to a method for predicting the quality of nutrient additives required to promote microbial growth based on a metabolic calculation model. Background Art

[0002] Genome-scale metabolic network models are mathematical representations of the entire cellular metabolic network of a given organism. These models are based on annotated genomes of the corresponding organism data, biochemical and organism-specific physiological information. These models describe the stoichiometry of metabolic networks, where metabolites are nodes and chemical reactions are edges. By defining mathematical constraints under which the metabolic network needs to operate, these models can be used to predict the growth of microorganisms or cell lines under different conditions, including various nutritional additives (i.e., chemical compounds, metabolites, such as amino acids). The basic mathematical method to perform such analysis is based on flux balance analysis (FBA). FBA is used to calculate the flux of metabolites through the metabolic network, the growth rate of a given organism, and the production rate of important metabolites (Orth et al., 2010). In addition, by simulating constraint-based stoichiometric genome-scale metabolic network models, one can predict which nutritional additives are essential for growth and which nutritional additives can affect the growth rate. For example, a study in Escherichia coli demonstrated how its genome-scale metabolic network model can be used as a platform to test the effects of various nutrient additives on improving growth (Marinos et al., 2020). The researchers measured the effectiveness of nutrient additives by the increase in biomass and found that carbohydrates had the strongest effect on growth yield, followed by oxygen and certain amino acids.

[0003] Although genome-scale metabolic network models can be used to predict beneficial growth nutritional additives, these analyses are likely to result in a large proportion of false positive results, where the predicted growth-beneficial nutritional additives show no or only minor improvements in actual cell culture. One factor contributing to this discrepancy is that genome-scale metabolic network models do not take into account other factors that affect cellular metabolism, such as the catalytic activity and molecular weight (MW) of enzymes. For example, the results of the aforementioned study (Marinos et al., 2020) were based solely on genome-scale metabolic network models and did not consider enzyme data as metabolic constraints. Incorporating enzyme data into metabolic models can match metabolic predictions to a biologically feasible range, making these models a more reliable platform for a variety of applications (such as the prediction of targeted growth supplements).

[0004] Genome-scale metabolic network models describe only the stoichiometry of metabolic networks, as they do not take into account various additional factors that have a major impact on the characteristics and behavior of cellular metabolism. Important factors, for example, are the catalytic activity of metabolic enzymes, and the cellular resources required to make these enzymes available (since cells have only a limited number of proteins that should ideally be allocated in an efficient manner). Enzyme catalytic activity is taken into account in kinetic metabolic models, which describe metabolic pathways via a set of ordinary differential equations; however, these model formulas typically do not cover genome-wide metabolism, but only describe a small part of the entire metabolic network that includes a small set of enzymes. Also, because genome-scale metabolic network models do not include enzyme kinetic data, they cannot make the most reliable predictions about the nutritional additives needed to promote growth.

[0005] It should be noted that the information disclosed in the above background technology section is only used for understanding the background of the present application, and therefore may include information that does not constitute prior art known to ordinary technicians in the field. Summary of the invention

[0006] The main purpose of the present invention is to overcome the defects existing in the above-mentioned background technology and provide a method for predicting the quality of nutritional additives required to promote the growth of microorganisms based on a metabolic calculation model.

[0007] To achieve the above object, the present invention adopts the following technical solutions:

[0008] A method for predicting the quality of nutritional additives required to promote microbial growth based on a metabolic computational model comprises the following steps:

[0009] S1. Obtain a known genome-scale metabolic network model of the target microorganism; or, perform genome analysis on the target microorganism, obtain the gene information in its genome, and construct a genome-scale metabolic network model including metabolic reactions and metabolites; the model is used to describe the metabolic network in microbial cells and calculate the flow of metabolites and the growth rate of microbial cells;

[0010] S2, collecting the molecular weight MW of the enzymes in the model and the turnover number of the enzyme-catalyzed reaction, i.e., the kcat value, setting the total cell protein concentration and the ratio of the metabolic protein in the total cell protein concentration, and setting the reaction rate limit based on the enzyme kinetic parameters, thereby introducing enzyme kinetics and protein pool data constraints in the genome-scale metabolic network model;

[0011] S3. Using the constrained genome-scale metabolic network model, by setting the exchange reactions of different nutritional additives and their uptake rates, the supplement specific efficiency SSE index of the target microorganism is calculated, and the maximum growth rate of the target microorganism after adding the nutritional additives is predicted, thereby evaluating the efficiency of the nutritional additives in improving the growth rate of the target microorganism.

[0012] In some optional embodiments, in step S1, the construction of the genome-scale metabolic network model includes: establishing an S matrix of metabolites and chemical reactions, which contains the stoichiometric coefficients of the compounds in the reactions, listing the compounds in the rows of the matrix, and listing the reactions in the columns of the matrix; defining the lower limit and upper limit of the reaction as separate vectors, selecting a reaction to be maximized or minimized as the objective function, and performing linear programming to calculate the flux and growth rate of the metabolites.

[0013] In some optional embodiments, the genome-scale metabolic network model realizes gene-protein-reaction relationships based on organism-specific genome and gene annotation information to reveal the association between enzymes determined by protein-coding genes and biochemical reactions.

[0014] In some optional embodiments, in step S1, the construction of the genome-scale metabolic network model further comprises:

[0015] Determine the macromolecular composition of cellular biomass and define it as a biomass pseudo-reaction;

[0016] Perform reaction fills to ensure biomass can be generated from a known culture medium;

[0017] The maximum growth rate of the model was determined by linear programming-based FBA, in which the biomass pseudo-reaction was set as the objective function to be maximized;

[0018] Manual calibration was performed to ensure that the model could predict the assimilation of known carbon and nitrogen sources, as well as the production of organic acids and secondary metabolites;

[0019] Introduce the ATP data required for maintaining growth-related and non-growth-related ATP into the model;

[0020] If the cells of the organism under study contain organelles, the compounds are assigned to their subcellular compartments based on the subcellular localization data of the enzymes;

[0021] Gap filling was performed to ensure that the metabolic network maintained connectivity after partitioning as before partitioning;

[0022] The model was calibrated by manual correction to ensure that the predicted growth rates and yields of known compounds were consistent with experimental observations.

[0023] In some optional embodiments, in step S2, the introduction of enzyme kinetics and protein pool data constraints specifically includes:

[0024] Collect the molecular weight and kcat value of each enzyme in the model, as well as the total cellular protein concentration and the proportion of metabolized proteins in the total cellular protein concentration;

[0025] The reaction rate was set not to exceed the product of the kcat value and the enzyme concentration, and the mass consumption of the enzyme was calculated by multiplying the reaction rate by the MW and then dividing by the kcat value;

[0026] Introducing enzyme-specific capacity and efficiency constraints to limit the reaction rate by using enzymes as entities in the model;

[0027] In the model, the stoichiometric constraint on the amount of a specific enzyme used is achieved by defining the proportional relationship between the kcat value and the MW of the enzyme, that is, the ratio of the MW to the kcat value.

[0028] In some optional implementations, step S2 includes:

[0029] Collect the molecular weight MW and kcat values ​​of each enzyme in the model from existing databases or by manual determination;

[0030] Estimate the proportion of total cellular protein concentration to biomass for a particular organism and the proportion of metabolic proteins in total cellular protein concentration that appear in genome-scale metabolic network models;

[0031] Estimate overall enzyme saturation factors for specific organisms;

[0032] The kcat value of a specific enzyme is obtained from known databases or scientific literature, or is directly measured by enzyme activity assay experiments using cell lysates of the target organism, or is estimated by a machine learning algorithm.

[0033] In some optional implementations, step S2 further includes:

[0034] Use the collected enzyme-specific kcat values ​​and MW to provide enzyme and reaction specific constraints for the model, where the maximum flux of each enzyme-catalyzed reaction is limited by the ratio of MW to kcat value;

[0035] The flux for each reaction was determined by multiplying the ratio of the MW to the kcat value, representing the amount of specific enzyme used, with the upper limit of the specific enzyme used being limited by the measured enzyme concentration.

[0036] In some optional embodiments, step S2 further includes: the total amount of all enzymes used is constrained by the protein pool of the total cellular protein concentration, that is, the product of its mass fraction and the mass fraction of the metabolized protein and the enzyme saturation factor.

[0037] In some optional implementations, step S2 further includes:

[0038] Testing whether genome-scale metabolic network models are consistent with actual phenotypes;

[0039] If the model predictions failed to reproduce the actual phenotypic results, the parameters were adjusted, especially the enzyme saturation factor and the enzyme-specific kcat value.

[0040] In some optional embodiments, step S2 further includes: if the phenotype predicted by the model is more efficient than the actual observed phenotype, reducing the saturation factor so that the phenotype predicted by the model is consistent with the actual observed phenotype result;

[0041] In some optional embodiments, step S2 also includes: using the genome-scale metabolic network model to perform flux balance analysis and metabolic network simulation, and after optimizing objective functions such as growth, performing another round of optimization, fixing the previous objective function value to minimize the usage of the total protein pool.

[0042] In some optional implementations, step S3 includes:

[0043] Set the maximum uptake rate of carbon source;

[0044] The maximum growth rate under the condition of adding specific nutrient additives was predicted through linear programming optimization model;

[0045] For each nutritional additive, an exchange reaction was added to the model to simulate the uptake of the nutritional additive by the organism;

[0046] Calculate the supplement specific efficiency (SSE) index to quantify the contribution of nutritional additives to improving growth rate;

[0047] Analyze model predictions to identify nutritional additives that significantly promote microbial growth.

[0048] In some optional embodiments, the carbon source is one or more of D-glucose, D-galactose, D-fructose, D-xylose, lactose, maltose, cellobiose, cellulose, and starch.

[0049] In some optional embodiments, the supplement specific efficiency SSE index is calculated by the following formula:

[0050]

[0051] Among them, μ max,spt is the maximum growth rate when nutritional supplements are provided, μ max is the maximum growth rate when no nutritional supplement is provided, q spt is the nutrient additive consumption rate, MW spt is the molecular weight of a potential nutritional additive.

[0052] In some optional implementations, step S3 includes:

[0053] Analyses were performed using genome-scale metabolic network models constrained by enzyme kinetics and protein pool data;

[0054] The maximum uptake rate of the carbon source can be set to an arbitrary value by setting the limits of the corresponding exchange reactions;

[0055] The growth rate was set as the objective function, and the model was simulated through flux balance analysis to predict the maximum growth rate.

[0056] In some optional implementations, step S3 further includes:

[0057] According to the results of flux balance analysis, the predicted uptake rate of various nutritional additives is determined, and the lower and upper limits of the exchange reaction of each nutritional additive in the model are set accordingly;

[0058] The maximum growth rate was fixed by setting the lower and upper limits of the biomass pseudo-response to the predicted maximum growth rate values.

[0059] In some optional implementations, step S3 also includes:

[0060] Predicting the minimum uptake rate of the carbon source, and setting the maximum uptake rate of the carbon source to the predicted minimum uptake rate by setting the boundary of the corresponding exchange reaction;

[0061] The upper limit of the biomass pseudoreaction was set to the maximum value in the model constrained by the enzyme kinetics and protein pool data.

[0062] In some optional implementations, step S3 also includes:

[0063] The effects of ingestion of each nutritional supplement on growth rate in the model were tested one by one;

[0064] For each nutrient additive, an exchange reaction was added to the model to simulate the uptake of that nutrient additive by the organism.

[0065] In some optional implementations, step S3 further includes:

[0066] The uptake rates of the tested nutrient additives were set to the same values ​​as the arbitrary values ​​initially set for the carbon source by setting limits for the corresponding exchange reactions;

[0067] Uptake of non-test nutrient additives was prevented by setting the limits of the corresponding exchange reactions to 0;

[0068] Prediction of maximum growth rates, where the maximum growth rate of a genome-scale metabolic network model was determined by a linear programming-based flux balance analysis with biomass pseudoreaction as the maximization objective function, the model being constrained by enzyme and protein pool data and incorporating supplement exchange reactions.

[0069] In some optional embodiments, the method further comprises the following steps:

[0070] S4. Based on the prediction results of the genome-scale metabolic network model and the experimental observation data, the Pearson linear correlation coefficient is calculated to quantify the strength and direction of the linear relationship between the model prediction value and the experimental observation value, so as to evaluate the correlation between the variables and assist in predictive analysis and data interpretation.

[0071] In some optional embodiments, the method further comprises the following steps:

[0072] S5. Perform a t-test on the calculated Pearson linear correlation coefficient to determine whether the observed correlation coefficient is statistically significant. Verify the significance of the correlation coefficient by evaluating the probability of observing the data when the null hypothesis is true.

[0073] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the method for predicting the quality of nutrient additives required to promote microbial growth based on a metabolic calculation model.

[0074] A computer program product comprises a computer program, wherein when the computer program is executed by a processor, the method for predicting the quality of nutrient additives required for promoting microbial growth based on a metabolic calculation model is implemented.

[0075] The present invention has the following beneficial effects:

[0076] The present invention provides an innovative method for predicting and evaluating the effects of nutritional additives on microbial growth rates. The method significantly improves the accuracy of prediction by constructing and optimizing a genome-scale metabolic network model and introducing enzyme kinetics and protein pool data constraints. Compared with traditional models, the model of the present invention can more accurately reflect the metabolic activities within microbial cells, thereby more reliably predicting which nutritional additives are essential for microbial growth. By calculating the supplement specific efficiency (SSE) index, the present invention can quantify the contribution of nutritional additives to increasing growth rate and identify nutritional additives that have a significant promoting effect on microbial growth. In addition, the present invention experimentally verifies the accuracy of model predictions and proves its superiority in predicting the efficiency of nutritional additives. These advantages make the present invention have important practical value in the optimization of microbial fermentation processes and biotechnology applications.

[0077] Other beneficial effects of the embodiments of the present invention will be further described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 It is a schematic structural diagram of an embodiment of the present invention.

[0079] Figure 2 Correlation between experimental and predicted values ​​for the effect of nutrient additives on growth efficiency, where the equation represents a linear function of the trend line. DETAILED DESCRIPTION

[0080] The following is a detailed description of the embodiments of the present invention. It should be emphasized that the following description is only exemplary and is not intended to limit the scope and application of the present invention.

[0081] In the existing technology, researchers can use genome-scale metabolic network models to predict which nutritional additives are essential for growth and which nutritional additives can affect growth rate. However, because genome-scale metabolic network models do not contain enzyme kinetic data, they cannot make the most reliable predictions for growth nutritional additives.

[0082] The present invention provides a prediction method for predicting the effect of nutritional additives on growth efficiency, which uses a genome-scale metabolic network model constrained by enzyme kinetics and protein pool data for prediction. The present invention introduces enzyme and protein pool data constraints to the genome-scale metabolic network model, so that the new model can more accurately predict the effect of a single nutritional additive on a cell phenotype (such as maximum growth rate). At the same time, the present invention uses the SSE metric to predict and rank the effects of multiple potential nutritional additives on the phenotype.

[0083] See also Figure 1 The embodiment of the present invention provides a method for predicting the quality of nutritional additives required to promote microbial growth based on a metabolic calculation model, comprising the following steps:

[0084] S1. Obtain a known genome-scale metabolic network model of the target microorganism; or, perform genome analysis on the target microorganism, obtain the gene information in its genome, and construct a genome-scale metabolic network model including metabolic reactions and metabolites; the model is used to describe the metabolic network in microbial cells and calculate the flow of metabolites and the growth rate of microbial cells;

[0085] S2, collecting the molecular weight MW of the enzymes in the model and the turnover number of the enzyme-catalyzed reaction, i.e., the kcat value, setting the total cell protein concentration and the ratio of the metabolic protein in the total cell protein concentration, and setting the reaction rate limit based on the enzyme kinetic parameters, thereby introducing enzyme kinetics and protein pool data constraints in the genome-scale metabolic network model;

[0086] S3. Using the constrained genome-scale metabolic network model, by setting the exchange reactions of different nutritional additives and their uptake rates, the supplement specific efficiency SSE index of the target microorganism is calculated, and the maximum growth rate of the target microorganism after adding the nutritional additives is predicted, thereby evaluating the efficiency of the nutritional additives in improving the growth rate of the target microorganism.

[0087] The method of the present invention can be used to predict the effectiveness of nutritional additives in improving the growth rate of microorganisms. Compared with the traditional genome-scale metabolic network model, the genome-scale metabolic network model constrained by enzyme kinetics and protein pool data in the present invention can more accurately predict the effect of nutritional additives on growth efficiency. The embodiment of the present invention uses the SSE metric on the genome-scale metabolic network model constrained by enzyme kinetics and protein pool data to predict the effectiveness of nutritional additives on microbial growth.

[0088] The specific embodiments of the present invention are further described below.

[0089] A computational method for effectively predicting the effect of nutritional additives on cell growth comprises the following steps:

[0090] 1. Construction of genome-scale metabolic network model:

[0091] 1.1. A genome-scale metabolic network model is a set of metabolic processes that determine the physiological and biochemical capabilities within the cells of a specific organism. The model assumes that the metabolic reaction rates are in a steady state. In mathematical terms, the genome-scale metabolic network model is converted into an S matrix, which contains the stoichiometric coefficients of the compounds in the reaction, with the compounds listed in the rows and the reactions listed in the columns. At the same time, the lower and upper limits of the reactions are defined as separate vectors. A reaction that should be maximized or minimized is selected as the objective function, and then a mathematical optimization called linear programming is performed. The output of the linear programming optimization is a flux vector, which shows the flux of all reactions that, as a set, contribute to the maximization or minimization of the flux of the target reaction. The model is based on organism-specific genome and gene annotation information, as well as the enzymes contained in the genome. In the genome-scale metabolic network model, this information is implemented in the form of gene-protein-reaction relationships, revealing the association between enzymes and biochemical reactions determined by protein-coding genes. Reactions have adjustable flux lower and upper limits to ensure the irreversibility of the reaction and can set a specific flux value range. Literature analysis based on a collection of these reactions can determine which compounds can be assimilated and / or produced by cells, which are called exchange reactions.

[0092] 1.2. Determine the macromolecular composition of the cell biomass and define it as a biomass pseudo reaction. If there is no relevant data for the organism, the corresponding data of phylogenetically close organisms can be used, for example, by phylogenetic tree analysis or biological macromolecular sequence comparison, such as 16S rRNA / 18S rRNA sequence analysis, whole genome sequence alignment, etc., to determine phylogenetically close organisms.

[0093] 1.3. Perform reaction fills to ensure that biomass can be generated from a known culture medium. This may require adding spontaneous reactions that do not have gene-protein-reaction rules in order to generate all known biomass components.

[0094] 1.4. The maximum growth rate of the genome-scale metabolic network model was determined by linear programming-based FBA, in which the biomass pseudo-reaction was set as the objective function to be maximized.

[0095] 1.5. Manual correction based on literature knowledge was performed to ensure that the genome-scale metabolic network model can predict the assimilation of known carbon and nitrogen sources, as well as the production of reported compounds such as organic acids and secondary metabolites.

[0096] 1.6. The ATP data required to maintain growth-related and non-growth-related functions are introduced into the model. The growth-related part is directly implemented in the biomass pseudo-reaction in the form of ATP hydrolysis, while the non-growth-related part is implemented as an independent ATP hydrolysis reaction with a fixed flux value.

[0097] 1.7. If the cells of the organism under study contain organelles, assign the compounds to their subcellular compartments based on subcellular localization data of enzymes associated with these compounds. Such data can be obtained from the literature or using computational tools such as DeepLoc and CELLO.

[0098] 1.8. Since the partitioning in the previous step will introduce gaps in the genome-scale metabolic network model, additional gap filling is required to ensure that the metabolic network maintains the connectivity before partitioning after partitioning.

[0099] 1.9. The model was calibrated by manual correction to ensure that the predicted growth rates and yields of known compounds were realistic.

[0100] 2. Construction of genome-scale metabolic network models constrained by enzyme kinetics and protein pool data:

[0101] 2.1. As input, reconstruction of genome-scale metabolic network models based on enzyme kinetics and protein pool data requires a high-quality genome-scale metabolic network model, which can be constructed according to step 1. Alternatively, genome-scale metabolic network models can also be obtained from scientific literature or databases such as BiGG and BioModels. The construction of genome-scale metabolic network models based on enzyme kinetics and protein pool data is based on the following inequality:

[0102] v≤k cat ·e,

[0103] Where v represents the enzyme reaction rate, k cat is the turnover number of the enzyme catalyzing the reaction, and e is the concentration of the enzyme. The minimum amount of enzyme required for the reaction is the reaction rate v divided by k cat , so the mass consumption of the enzyme is the rate v multiplied by the molecular weight (MW) of the enzyme divided by k cat These enzyme-specific capacity and efficiency constraints can be computationally applied by treating the corresponding enzyme as an entity in the model (e.g., a metabolite) and therefore as a substrate for the reaction. In the model, the consumption of this pseudo-substrate represents the usage of the enzyme in catalyzing the corresponding reaction, and the stoichiometry of the reaction-specific enzyme usage can be determined by defining the enzyme's k cat and MW value, that is, MW and k cat Ratio.

[0104] 2.2 Collect information about each enzyme in the model. For example, the molecular weight (MW) value can be obtained by obtaining proteome data from other databases such as UniProt, KEGG, MetaCyc, or by manual determination.

[0105] 2.3. Estimate the total cellular protein concentration as a percentage of biomass (grams of protein / grams of dry cell weight) for a specific organism directly from experiments or literature; or assume this ratio to be 0.5 as described in the literature (Nilsson & Nielsen, 2016).

[0106] 2.4. Estimate the proportion of metabolic proteins in the total cellular protein concentration that appear in the genome-scale metabolic network model directly from experiments or literature; or assume that the proportion is 0.5 as described in the literature (Nilsson & Nielsen, 2016).

[0107] 2.5. Estimate the overall enzyme saturation factor for the specific organism from experiments or literature; or assume it to be 0.5 as described in the literature (Nilsson & Nielsen, 2016).

[0108] 2.6. Obtain the k value of a specific enzyme from BRENDA, SABIO-RK or other databases cat value.

[0109] 2.7. Alternatively, the k values ​​for specific enzymes can be manually collected from the scientific literature. cat value.

[0110] 2.8. Alternatively, the k of a specific enzyme can be measured directly by performing enzyme activity assays on cell lysates of the target organism. cat value.

[0111] 2.9. Alternatively, the k of a specific enzyme can be estimated by machine learning algorithms such as PreKcat, DeepEnzyme, or DLKcat. cat value.

[0112] 2.10. Collected enzyme-specific k cat The MW values ​​are used to provide enzyme and reaction specificity constraints for the model. For each enzyme-catalyzed reaction, the maximum flux of the reaction is given by the relationship between MW and k cat limited by the ratio.

[0113] 2.11. Multiply the flux of each reaction by MW and k cat The ratio of , represents the amount of a specific enzyme used. The upper limit of the amount of a specific enzyme used can be limited by the measured enzyme concentration, which can be obtained through proteomic quantification experiments.

[0114] 2.12. The total amount of all enzymes used is constrained by the protein pool of the total cellular protein concentration, that is, the product of its mass fraction and the mass fraction of metabolized proteins and the enzyme saturation factor.

[0115] 2.13. Genome-scale metabolic network models constrained by enzyme kinetics and protein pool data will be tested for consistency with phenotypes (e.g., maximum growth rate). If model predictions fail to reproduce actual phenotypic outcomes, some parameters, particularly enzyme saturation factors and enzyme specificity k, may be adjusted. cat value.

[0116] 2.14. If the model predicts a phenotype that is more efficient than what is observed, such as a higher maximum growth rate, the saturation factor can be reduced so that the model-predicted phenotype is consistent with the observed phenotype.

[0117] 2.15. Functional genome-scale metabolic network models based on enzyme and protein pool data can be used for flux balance analysis and other simulations, similar to genome-scale metabolic network models that are not constrained by enzyme kinetics and protein pool data. After optimizing an objective function such as growth, another round of optimization can be performed to fix the previous objective function value and minimize the use of the total protein pool.

[0118] 3. Calculate the efficiency of nutrient additives in increasing growth rate:

[0119] 3.1. Genome-scale metabolic network models constrained by enzyme and protein pool data were used for this analysis.

[0120] 3.2. The maximum uptake rate of the carbon source can be set to an arbitrary value by setting the limits of the corresponding exchange reaction. 3.3. The carbon source can be D-glucose, D-galactose, D-fructose, D-xylose, lactose, maltose, cellobiose, cellulose, starch, etc.

[0121] 3.4. Set the growth rate as the objective function, simulate the model through flux balance analysis, and predict the maximum growth rate.

[0122] 3.5. The maximum growth rate derived from the genome-scale metabolic network model constrained by enzyme kinetics and protein pools was determined by a flux balance analysis based on linear programming, with the biomass pseudo-reaction as the maximization objective function, and the actual flux was achieved when the absolute sum of the fluxes was minimized.

[0123] 3.6. From the same flux balance analysis results, obtain predicted uptake rates for other nutritional additives based on the flux values ​​and use these values ​​to set lower and upper limits for the corresponding exchange reactions.

[0124] 3.7. Fix the maximum growth rate by setting the lower and upper limits of the biomass pseudo-response to the predicted maximum growth rate values.

[0125] 3.8 Prediction of the minimum uptake rate of carbon sources The minimum uptake rate of carbon sources for the genome-scale metabolic network model was determined by a linear programming-based flux balance analysis with carbon source exchange reactions as the minimization objective function, and the model was constrained by enzyme and protein pool data.

[0126] 3.9. The maximum uptake rate of the carbon source is set to the predicted minimum uptake rate by setting the boundaries of the corresponding exchange reactions.

[0127] 3.10. Upper limit of biomass pseudo-reactions to maximum in genome-scale metabolic network models constrained by enzyme kinetics and protein pool data.

[0128] 3.11. Test the effect of intake of each nutritional supplement on the growth rate in the model one by one.

[0129] 3.12. Nutritional additives can be any compound that appears in the genome-scale metabolic network model constrained by enzyme kinetics and protein pool data.

[0130] 3.13. For each nutritional additive, an exchange reaction was added to the genome-scale metabolic network model constrained by the enzyme and protein pool data to simulate the uptake of the nutritional additive by the organism.

[0131] 3.14. Set the absorbance of the tested nutrient additive to the same arbitrary value as originally set for the carbon source by setting the limits of the corresponding exchange reaction.

[0132] 3.15. Prevent uptake of other tested nutritional additives by setting the limits of the corresponding exchange reactions to 0.

[0133] 3.16. Predict the maximum growth rate.

[0134] 3.17. The maximum growth rate of the genome-scale metabolic network model was determined by a linear programming-based flux balance analysis with the biomass pseudoreaction as the maximization objective function. The model was constrained by enzyme and protein pool data and included supplement exchange reactions.

[0135] 3.18. Calculate the supplement specific efficiency (SSE) index using the following formula:

[0136]

[0137] Among them, μ max,spt is the maximum growth rate when nutrient supplementation is provided (g newly produced cell dry weight / g (existing) cell dry weight / hour), μ max is the maximum growth rate when no nutritional supplements are provided (g newly produced cell dry weight / g (existing) cell dry weight / hour), q spt is the rate of nutrient supplement consumption (mmol nutrient supplement / g (existing) cell dry weight / hour), q spt,0 is the consumption rate of the nutritional supplement when no additional nutritional supplement is provided (mmol supplement / g (existing) cell dry weight / hour), MW spt is the molecular weight of the potential nutritional additive (grams of nutritional additive / mole of nutritional additive). In the formula, MW spt Divide by 1000 to convert the units to grams of nutrient additive / mmol of nutrient additive. spt,0 is always zero, the equation can be simplified to the following formula:

[0138]

[0139] The unit of SSE is gram of newly produced cell dry weight / gram of nutrient supplement.

[0140] 4. Pearson linear correlation coefficient:

[0141] 4.1. Pearson's linear correlation coefficient, denoted as r, is a statistical measure used to quantify the strength and direction of the linear relationship between two continuous variables. It is calculated as the covariance of the two variables divided by the product of their standard deviations. The coefficient ranges from -1 to 1, where r = 1 indicates a perfect positive linear relationship, r = -1 indicates a perfect negative linear relationship, and r = 0 indicates no linear relationship. This coefficient is widely used in various fields to assess the degree of correlation between variables, which helps in predictive analysis and data interpretation.

[0142] 4.2. The formula for Pearson's linear correlation coefficient is:

[0143]

[0144] where x i and i is a single sample point, and are the means of the x and y variables respectively and n is the sample size.

[0145] 5. t-test for linear correlation:

[0146] 5.1 To determine whether a linear correlation between two sets of data is statistically significant, a t-test can be performed. The t-test assesses whether the observed correlation coefficient r is significantly different from zero (no correlation). The test statistic t is calculated using the following formula:

[0147]

[0148] Where r is the Pearson correlation coefficient and n is the number of data points. The degrees of freedom for this test are n-2. The p-value is then drawn from a t-distribution with n-2 degrees of freedom. It represents the probability of observing the data assuming that the null hypothesis (no correlation) is true. A low p-value (usually < 0.05) indicates that the correlation is statistically significant, meaning that the probability that the observed correlation occurred by chance is low.

[0149] Example 1: Reconstruction of a genome-scale metabolic network model of Aspergillus oryzae constrained by enzyme kinetics and protein pool data

[0150] The publicly available A. oryzae genome-scale metabolic network model version iWV1314 was used. The iWV1314 model was modified to take into account the utilization of enzymes. Enzymes were added as pseudosubstrates in the model and the corresponding reactions were constrained by using MW-to-cat ratios. Enzyme-specific molecular weight (MW) values ​​were derived from the UniProt proteome UP000006564, k cat The values ​​were taken from the BRENDA database. For specific enzymes whose reaction k values ​​were not available from BRENDA, catIf the value is cat Predicting it, the specific steps may include: determining the identity of the enzyme; accessing the DLKcat tool or database; inputting the information of the enzyme; running the prediction; obtaining the prediction results; evaluating the accuracy of the prediction; integrating into the model. The total protein content is set to 0.2-0.8, for example, 0.5 grams of protein / gram of cell dry weight. The mass fraction of metabolic proteins in the total protein is calculated from the proteomics data collected in the PRIDE database and set to 0.2-0.8, for example, 0.38. The enzyme saturation factor is set to 0.2-0.8, for example, 0.5.

[0151] Since the maximum growth rate calculated by the model was 0.12 / h, which was significantly lower than the highest growth rate reported for A. oryzae (0.38 / h), the enzyme-specific k values ​​of eight enzymes and one enzyme complex were manually corrected. cat The value was adjusted to ensure that the maximum growth rate was close to 0.38 / hour. After manual correction, the maximum growth rate of the genome-scale metabolic network model constrained by enzyme kinetics and protein pool data reached 0.44 / hour. Table 1 provides the detailed information of the resulting enzyme-constrained genome-scale metabolic network model and the corresponding values ​​in iVW1314:

[0152] Table 1

[0153] category iWV1314 quantity Number of enzyme constraint models reconstructed reaction 2,361 4,173 Metabolites 1,104 2,403 Gene 1,346 1,347 Cell compartment 4 4

[0154] Table 2 lists the corrected k cat Enzyme list of values ​​(in h-1):

[0155] Table 2

[0156]

[0157] The minimum size of the predicted total cellular protein pool from a genome-scale metabolic network model constrained by enzyme kinetics and protein pool data was 124 mg enzyme protein per gram cell dry weight.

[0158] Example 2: Computing the effectiveness of nutrient additives on increasing growth rate using a genome-scale metabolic network model of Aspergillus oryzae constrained by enzyme kinetics and protein pool data

[0159] D-glucose was selected as the carbon source, and its maximum uptake rate was set to 1 mmol / g cell dry weight / hour by setting an upper limit for the D-glucose exchange reaction. The predicted maximum growth rate was 0.135 / hour. Based on the linear programming analysis, the following flux values ​​were used to limit the exchange reactions of other growth medium components:

[0160] Ammonium: 0.83 mmol / g cell dry weight / hour

[0161] Oxygen: 3.1 mmol / g cell dry weight / hour

[0162] Hydrogen phosphate: 0.09 mmol / g cell dry weight / hour

[0163] · Hydrogen sulfate: 0.01 mmol / g cell dry weight / hour

[0164] The growth rate was then fixed to a maximum by setting the lower and upper limits of the biomass pseudo-reaction to 0.135 mmol / g cell dry weight / hour. The predicted minimum D-glucose uptake rate was 0.99 mmol / g cell dry weight / hour, so the upper limit of the D-glucose exchange reaction was set to this value. The lower and upper limits of the biomass pseudo-reaction were changed to 0 mmol / g cell dry weight / hour and 1,000 mmol / g cell dry weight / hour, respectively.

[0165] The following seven amino acids were tested for their effects on cell growth: L-alanine, L-isoleucine, L-leucine, L-methionine, L-proline, L-valine, and glycine. The maximum growth rate was predicted and used to calculate the SSE value by setting the upper limit of the corresponding exchange reaction and setting its maximum absorbance to 1 mmol / g cell dry weight / hour. Before switching to the next tested amino acid, the maximum absorbance of the currently tested amino acid was set to 0 mmol / g cell dry weight / hour by setting the upper limit of the corresponding exchange reaction.

[0166] The calculated SSE (grams of newly produced cell dry weight / grams of nutrient additive) values ​​are shown in Table 3:

[0167] Table 3

[0168] Nutritional Additives SSE L-Alanine 0.00707 L-Isoleucine 0.00489 L-Leucine 0.00644 L-Methionine 0.00117 L-Proline 0.00886 L-Valine 0.00871 Glycine 0.01359

[0169] Example 3: Calculation of the effectiveness of nutritional additives in increasing growth rate using the Aspergillus oryzae genome-scale metabolic network model iVW1314 (without enzyme kinetics and protein pool data constraints).

[0170] D-glucose was selected as the carbon source, and its maximum uptake rate was set to 1 mmol / g cell dry weight / hour by setting an upper limit for the D-glucose exchange reaction. The predicted maximum growth rate was 0.138 / hour. Based on the linear programming analysis, the following flux values ​​were used to limit the exchange reactions of other growth medium components:

[0171] Ammonium: 0.84 mmol / g cell dry weight / hour

[0172] Oxygen: 3.2 mmol / g cell dry weight / hour

[0173] Hydrogen phosphate: 0.1 mmol / g cell dry weight / hour

[0174] · Hydrogen sulfate: 0.01 mmol / g cell dry weight / hour

[0175] The growth rate was then fixed to a maximum value by setting the lower and upper limits of the biomass pseudo-reaction to 0.138 mmol / g cell dry weight / hour. The predicted minimum D-glucose uptake rate was 0.99 mmol / g cell dry weight / hour, so the upper limit of the D-glucose exchange reaction was set to this value. The lower and upper limits of the biomass pseudo-reaction were changed to 0 mmol / g cell dry weight / hour and 1,000 mmol / g cell dry weight / hour, respectively.

[0176] The following seven amino acids were tested for their effects on cell growth: L-alanine, L-isoleucine, L-leucine, L-methionine, L-proline, L-valine, and glycine. The maximum growth rate was predicted and used to calculate the SSE value by setting the upper limit of the corresponding exchange reaction and setting its maximum absorbance to 1 mmol / g cell dry weight / hour. Before switching to the next tested amino acid, the maximum absorbance of the currently tested amino acid was set to 0 mmol / g cell dry weight / hour by setting the upper limit of the corresponding exchange reaction.

[0177] The calculated SSE (grams of newly produced cell dry weight / grams of nutrient additive) values ​​are shown in Table 4:

[0178] Table 4

[0179] Nutritional Additives SSE L-Alanine 0.00751 L-Isoleucine 0.00526 L-Leucine 0.00526 L-Methionine 0.00003 L-Proline 0.00599 L-Valine 0.00589 Glycine 0.00211

[0180] Example 4. Verification of the effects of seven amino acids on the growth of Aspergillus oryzae

[0181] In order to investigate whether the model with enzyme dynamics and protein pool constraints can predict the effectiveness of nutrient additives better than the original iVW1314 model, an Aspergillus oryzae strain was used for experimental verification, and the effects of seven amino acids (L-alanine, L-isoleucine, L-leucine, L-methionine, L-proline, L-valine, and glycine) on the biomass of Aspergillus oryzae were tested. The basal medium used for the experiment contained 10 g / L glucose, 5.95 g / L NaNO3, 0.52 g / L KCl, 1.52 g / L KH2PO4, 0.49 g / L MgSO4·7H2O, and 1 mL of trace element solution. The trace element solution contained 1gNa2EDTA, 0.1g FeSO4·7H2O, 0.14g CaCl2·2H2O, 0.1g MnCl2·4H2O, 0.02g Na2MoO4·2H2O, 0.03g CuSO4·5H2O, 0.032g CoCl2·6H2O, and 0.44gZnSO4·7H2O. The culture experiment was carried out in a 250mL shake flask, which was filled with 50mL of basal culture medium. One of the seven amino acids was added to each shake flask, and the concentration of the added amino acid was 50mmol / L. The shake flask without amino acid addition was used as a control. The inoculation concentration was 5×10 7 The spores / mL were cultured in a shake flask at 37°C and a stirring speed of 100 rpm for 24 hours, then filtered and freeze-dried, and the cell dry weight was measured and converted to the final mycelium concentration. The obtained Aspergillus oryzae biomass results are shown in Table 5:

[0182] Table 5

[0183]

[0184] Example 5: Comparison of the performance of different genome-scale metabolic network models in predicting the effect of nutritional supplements on strain growth efficiency

[0185] Comparing the SSE values ​​calculated in Examples 3 and 4 with the experimental data, we can obtain the correlation between the experimental and predicted values ​​( Figure 1). The results showed that there was a strong linear correlation between the experimental data and the predicted SSE values ​​of the genome-scale metabolic network model constrained by enzyme kinetics and protein pool data (r = 0.96), while the correlation between the experimental data and the predicted SSE values ​​of the genome-scale metabolic network model iVW1314 was weaker (r = 0.31). To evaluate the significance of the prediction, a linear correlation t-test was performed. The p-value obtained for the genome-scale metabolic network model constrained by enzyme kinetics and protein pool data was 0.0005, while the corresponding value for the genome-scale metabolic network model was 0.5. It can therefore be concluded that the genome-scale metabolic network model constrained by enzyme kinetics and protein pool data is more accurate than the genome-scale metabolic network model in predicting the efficiency of nutritional additives.

[0186] In summary, the present invention provides an innovative prediction method for predicting the effect of nutritional additives on growth efficiency, and uses a genome-scale metabolic network model constrained by enzyme kinetics and protein pool data to evaluate the effect of nutritional additives on microbial growth efficiency. This method significantly improves the accuracy of prediction and can more accurately simulate the metabolic activities in microbial cells compared with traditional models. Taking Aspergillus oryzae as an example, the model prediction of the present invention shows a strong linear correlation with the experimental results (r=0.96, p value=0.0005), which is much better than the traditional model (r=0.31, p value=0.5). By introducing enzyme and protein pool data constraints, the present invention can not only accurately predict the effect of a single nutritional additive, but also use the supplement specific efficiency (SSE) metric to rank and predict the effects of multiple potential nutritional additives, providing a powerful tool for the optimization of microbial fermentation processes and biotechnology applications.

[0187] An embodiment of the present invention further provides a storage medium for storing a computer program, which at least performs the above method when executed.

[0188] An embodiment of the present invention further provides a control device, comprising a processor and a storage medium for storing a computer program; wherein the processor is configured to execute at least the method described above when executing the computer program.

[0189] An embodiment of the present invention further provides a processor, wherein the processor executes a computer program and at least executes the method described above.

[0190] The storage medium can be implemented by any type of non-volatile storage device, or a combination thereof. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a ferromagnetic random access memory (FRAM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM); the magnetic surface memory can be a disk memory or a tape memory. The storage medium described in the embodiments of the present invention is intended to include, but is not limited to, these and any other suitable types of memory.

[0191] In the several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0192] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0193] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0194] Those skilled in the art can understand that: all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), disks or optical disks, etc. Various media that can store program codes.

[0195] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0196] The methods disclosed in the several method embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments.

[0197] The features disclosed in several product embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new product embodiments.

[0198] The features disclosed in several method or device embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

[0199] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art of the present invention, several equivalent substitutions or obvious variations can be made without departing from the concept of the present invention, and the performance or use is the same, which should be regarded as belonging to the protection scope of the present invention.

Claims

1. A method for predicting the quality of nutritional additives required to promote microbial growth based on a metabolic calculation model, characterized in that: The following steps are involved: S1. Obtain a known genome-scale metabolic network model of the target microorganism; or, perform genome analysis on the target microorganism, obtain the gene information in its genome, and construct a genome-scale metabolic network model including metabolic reactions and metabolites; the model is used to describe the metabolic network in microbial cells and calculate the flow of metabolites and the growth rate of microbial cells; S2, collecting the molecular weight MW of the enzymes in the model and the turnover number of the enzyme-catalyzed reaction, i.e., the kcat value, setting the total cell protein concentration and the ratio of the metabolic protein in the total cell protein concentration, and setting the reaction rate limit based on the enzyme kinetic parameters, thereby introducing enzyme kinetics and protein pool data constraints in the genome-scale metabolic network model; S3. Using the constrained genome-scale metabolic network model, by setting the exchange reactions of different nutritional additives and their uptake rates, the supplement specific efficiency SSE index of the target microorganism is calculated, and the maximum growth rate of the target microorganism after adding the nutritional additives is predicted, thereby evaluating the efficiency of the nutritional additives in improving the growth rate of the target microorganism.

2. The method according to claim 1, characterized in that In step S1, the construction of the genome-scale metabolic network model includes: establishing an S matrix of metabolites and chemical reactions, which contains the stoichiometric coefficients of the compounds in the reactions, listing the compounds in the rows of the matrix, and listing the reactions in the columns of the matrix; defining the lower limit and upper limit of the reaction as separate vectors, selecting a reaction that should be maximized or minimized as the objective function, and performing linear programming to calculate the flux and growth rate of the metabolites.

3. The method according to claim 2, characterized in that The genome-scale metabolic network model realizes gene-protein-reaction relationships based on organism-specific genome and gene annotation information to reveal the association between enzymes determined by protein-coding genes and biochemical reactions.

4. The method according to claim 3, characterized in that In step S1, the construction of the genome-scale metabolic network model further includes: Determine the macromolecular composition of cellular biomass and define it as a biomass pseudo-reaction; Perform reaction fills to ensure biomass can be generated from a known culture medium; The maximum growth rate of the model was determined by linear programming-based FBA, in which the biomass pseudo-reaction was set as the objective function to be maximized; Manual calibration was performed to ensure that the model could predict the assimilation of known carbon and nitrogen sources, as well as the production of organic acids and secondary metabolites; Introduce the ATP data required for maintaining growth-related and non-growth-related ATP into the model; If the cells of the organism under study contain organelles, the compounds are assigned to their subcellular compartments based on the subcellular localization data of the enzymes; Gap filling was performed to ensure that the metabolic network maintained connectivity after partitioning as before partitioning; The model was calibrated by manual correction to ensure that the predicted growth rates and yields of known compounds were consistent with experimental observations.

5. The method according to any one of claims 1 to 4, characterized in that: In step S2, the introduction of enzyme kinetics and protein pool data constraints specifically includes: Collect the molecular weight and kcat value of each enzyme in the model, as well as the total cellular protein concentration and the proportion of metabolized proteins in the total cellular protein concentration; The reaction rate was set not to exceed the product of the kcat value and the enzyme concentration, and the mass consumption of the enzyme was calculated by multiplying the reaction rate by the MW and then dividing by the kcat value; Introducing enzyme-specific capacity and efficiency constraints to limit the reaction rate by using enzymes as entities in the model; In the model, the stoichiometric constraint on the amount of a specific enzyme used is achieved by defining the proportional relationship between the kcat value and the MW of the enzyme, that is, the ratio of the MW to the kcat value.

6. The method according to any one of claims 1 to 4, characterized in that: Step S2 includes: Collect the molecular weight MW and kcat values ​​of each enzyme in the model from existing databases or by manual determination; Estimate the proportion of total cellular protein concentration to biomass for a particular organism and the proportion of metabolic proteins in total cellular protein concentration that appear in genome-scale metabolic network models; Estimate the overall enzyme saturation factor for a specific organism; The kcat value of a specific enzyme is obtained from known databases or scientific literature, or is directly measured by enzyme activity assay experiments using cell lysates of the target organism, or is estimated by a machine learning algorithm.

7. The method according to any one of claims 1 to 4, characterized in that Step S2 also includes: Use the collected enzyme-specific kcat values ​​and MW to provide enzyme and reaction specific constraints for the model, where the maximum flux of each enzyme-catalyzed reaction is limited by the ratio of MW to kcat value; The flux for each reaction was determined by multiplying the ratio of the MW to the kcat value, representing the amount of specific enzyme used, with the upper limit of the specific enzyme used being limited by the measured enzyme concentration.

8. The method according to any one of claims 1 to 4, characterized in that Step S2 also includes: the total amount of all enzymes used is constrained by the protein pool of the total cell protein concentration, that is, the product of its mass fraction and the mass fraction of the metabolized protein and the enzyme saturation factor.

9. The method according to any one of claims 1 to 4, characterized in that: Step S2 also includes: Testing whether genome-scale metabolic network models are consistent with actual phenotypes; If the model predictions failed to reproduce the actual phenotypic results, the parameters were adjusted, especially the enzyme saturation factor and the enzyme-specific kcat value.

10. The method according to claim 9, characterized in that Step S2 also includes: if the phenotype predicted by the model is more efficient than the actual observed phenotype, reducing the saturation factor to make the phenotype predicted by the model consistent with the actual observed phenotype result.

11. The method according to any one of claims 1 to 4, characterized in that: Step S2 also includes: using the genome-scale metabolic network model to perform flux balance analysis and metabolic network simulation, and after optimizing objective functions such as growth, performing another round of optimization, fixing the previous objective function value to minimize the usage of the total protein pool.

12. The method according to any one of claims 1 to 4, characterized in that: Step S3 includes: Set the maximum uptake rate of carbon source; The maximum growth rate under the condition of adding specific nutrient additives was predicted through linear programming optimization model; For each nutritional additive, an exchange reaction was added to the model to simulate the uptake of the nutritional additive by the organism; Calculate the supplement specific efficiency (SSE) index to quantify the contribution of nutritional additives to improving growth rate; Analyze model predictions to identify nutritional additives that significantly promote microbial growth.

13. The method according to claim 12, characterized in that The carbon source is one or more of D-glucose, D-galactose, D-fructose, D-xylose, lactose, maltose, cellobiose, cellulose and starch.

14. The method according to claim 12, characterized in that The supplement specific efficiency SSE index is calculated by the following formula: Among them, μ max,spt is the maximum growth rate when nutritional supplements are provided, μ max is the maximum growth rate when no nutritional supplement is provided, q spt is the nutrient additive consumption rate, MW spt is the molecular weight of a potential nutritional additive.

15. The method according to claim 12, characterized in that Step S3 includes: Analyses were performed using genome-scale metabolic network models constrained by enzyme kinetics and protein pool data; The maximum uptake rate of the carbon source can be set to an arbitrary value by setting the limits of the corresponding exchange reactions; The growth rate was set as the objective function, and the model was simulated through flux balance analysis to predict the maximum growth rate.

16. The method according to claim 15, characterized in that Step S3 also includes: According to the results of flux balance analysis, the predicted uptake rate of various nutritional additives is determined, and the lower and upper limits of the exchange reaction of each nutritional additive in the model are set accordingly; The maximum growth rate was fixed by setting the lower and upper limits of the biomass pseudo-response to the predicted maximum growth rate values.

17. The method according to claim 12, characterized in that Step S3 also includes: Predicting the minimum uptake rate of the carbon source, and setting the maximum uptake rate of the carbon source to the predicted minimum uptake rate by setting the boundary of the corresponding exchange reaction; The upper limit of the biomass pseudoreaction was set to the maximum value in the model constrained by the enzyme kinetics and protein pool data.

18. The method according to claim 12, characterized in that Step S3 also includes: The effects of ingestion of each nutritional supplement on growth rate in the model were tested one by one; For each nutrient additive, an exchange reaction was added to the model to simulate the uptake of that nutrient additive by the organism.

19. The method of claim 12, wherein: Step S3 also includes: The uptake rates of the tested nutrient additives were set to the same values ​​as the arbitrary values ​​initially set for the carbon source by setting limits for the corresponding exchange reactions; Uptake of non-test nutrient additives was prevented by setting the limits of the corresponding exchange reactions to 0; Prediction of maximum growth rates, where the maximum growth rate of a genome-scale metabolic network model was determined by a linear programming-based flux balance analysis with biomass pseudoreaction as the maximization objective function, the model being constrained by enzyme and protein pool data and incorporating supplement exchange reactions.

20. The method according to any one of claims 1 to 19, characterized in that The following steps are also included: S4. Based on the prediction results of the genome-scale metabolic network model and the experimental observation data, the Pearson linear correlation coefficient is calculated to quantify the strength and direction of the linear relationship between the model prediction value and the experimental observation value, so as to evaluate the correlation between the variables and assist in predictive analysis and data interpretation.

21. The method of claim 20, wherein: The following steps are also included: S5. Perform a t-test on the calculated Pearson linear correlation coefficient to determine whether the observed correlation coefficient is statistically significant. Verify the significance of the correlation coefficient by evaluating the probability of observing the data when the null hypothesis is true.

22. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for predicting the quality of nutritional additives required to promote microbial growth based on a metabolic calculation model as described in any one of claims 1 to 21 is implemented.

23. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method for predicting the quality of nutritional additives required to promote microbial growth based on a metabolic calculation model as described in any one of claims 1 to 21 is implemented.

Images