A method and system for developing a feed medium based on metabolic parameters

Abstract

The application discloses a method and system for developing a feeding medium based on metabolic parameters. The method comprises the following steps: (1) metabolic parameter fitting: for a selected basic medium formula and cell culture target parameters, a metabolic parameter fitting model is used to fit metabolic parameters of cell culture to a preset time t0; (2) feeding medium development: according to the metabolic parameters of cell culture to the preset time obtained in step (1) and the cell state at the feeding medium addition time, metabolic consumption of each medium component at the feeding medium addition time is estimated; the concentration of each component of the feeding medium is determined, so that the feeding medium component used at this time meets the metabolic consumption of each medium component at this time. The regression model can greatly shorten the development time of different cell formulas, reduce the learning cost of talents, avoid the emergence of poor formulas due to lack of knowledge, improve the accuracy of the test, and avoid waste.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of biotechnology, and more particularly relates to a method and system for developing a feed medium based on metabolic parameters. BACKGROUND

[0002] Although the nutritional requirements of various microorganisms, cells and other biological samples are not the same, the basic nutrients required by most biological samples are the same. Currently, some intelligent development methods for cell basic medium have appeared, for example, Chinese patent document CN113450882A provides a basic medium formula development method and system based on artificial intelligence, applies machine learning algorithm to complex formula optimization process, builds a sample formula database with good quality and sufficient quantity, selects appropriate machine learning algorithm and optimization algorithm, and recommends the basic medium formula with the best culture effect in a short time to reduce the formula development threshold. For example, Chinese patent document CN113450868A, as an improvement of the former, further improves the culture effect of the developed basic medium.

[0003] However, the intelligent development method of the medium is only for the basic medium. In the case that the number of prior formula database is limited, or the cell culture period is long and the culture process is complex, the culture effect is not as good as the cell culture process combining the basic medium and the feed medium by manual exploration. SUMMARY

[0004] In view of the above defects or improvement needs of the prior art, the present application provides a method and system for developing a feed medium based on metabolic parameters, which aims to estimate the consumption of each component of the basic medium caused by cell metabolism when adding the feed medium, thereby replacing the complex process of manually exploring the feed medium, and thereby solving the technical problems of long development period and high professional threshold of the prior art for developing the feed medium.

[0005] To achieve the above-mentioned purpose, according to one aspect of the present application, a method for developing a feed medium based on metabolic parameters is provided, which comprises the following steps:

[0006] (1) Metabolic parameter fitting: for the selected basic medium formula and the cell culture target parameters, a metabolic parameter fitting model is used to fit the metabolic parameters of cell culture to a preset time t0 The metabolic parameters are the consumptions of each component in the basic medium when the cell culture is carried out according to the formula of the basic medium to a preset time;

[0007] (2) Feed medium development: according to the metabolic parameters of the cells cultured to the preset time obtained in step (1), and the cell state at the feed medium addition time, estimate the metabolic consumption amount of each medium component at the feed medium addition time; determine the concentration of each component of the feed medium, so that the feed medium component used at this time meets the metabolic consumption amount of each medium component at this time.

[0008] Preferably, the feed medium development method based on metabolic parameters, the cell culture target parameter is the target cell state required to be reached by cell culture, and the cell state is cell viability, cell density, and / or a biochemical indicator, and the biochemical indicator is protein expression amount, glucose, lactic acid, ammonia, and / or glutamine content.

[0009] Preferably, the feed medium development method based on metabolic parameters, the metabolic parameter fitting model inputs the basic medium formula and the cell culture target parameter; and the output is the metabolic parameter.

[0010] Preferably, the feed medium development method based on metabolic parameters, the metabolic parameter fitting model includes but is not limited to: a linear regression model, a support vector machine regression model, a K-nearest neighbor model, XGBoost, ridge regression, LightGBM, random forest, GBDT, or a deep learning model; and the deep learning model includes but is not limited to: a fully connected neural network, a convolutional neural network, a recurrent neural network, or an attention model.

[0011] Preferably, the feed medium development method based on metabolic parameters, the cell culture target parameter is a biochemical indicator, and the metabolic parameter fitting model is a linear model.

[0012] Preferably, the feed medium development method based on metabolic parameters, the metabolic parameter fitting model is obtained according to the following method:

[0013] (1-1) Collect training formulas;

[0014] (1-2) Obtain metabolic parameter data:

[0015] Using the basic medium formula formed in step (1-1), cell culture experiments are carried out to a preset time t0, the remaining concentrations of each component of the training formula obtained in step (1-1) are obtained, and the difference between the starting concentration and the remaining concentration of each component of the medium is calculated as the metabolic parameter, i.e. the consumption data of each component of the basic medium;

[0016] (1-3) Organize training sample data set:

[0017] The training sample data set is organized by taking the training formula ingredient addition amount obtained in step (1-1) and the original data or normalized values of the cell culture target parameters as the input matrix, and taking the metabolic parameter data obtained in step (1-2) as the output matrix.

[0018] (1-4) Training the fitting model with the training sample data set obtained in step (1-3) to obtain the metabolic parameter fitting model.

[0019] Preferably, the metabolic parameter-based feed medium development method adopts a function Estimating the metabolic consumption amount of each medium ingredient at the feed medium addition time t Where t = 0, f(t) = [0, 0,...], that is, when the culture time is 0, the consumption amount of each ingredient of the medium is also 0; t = t0, That is, when the culture time is preset time t0, the consumption amount of each ingredient of the medium is the metabolic parameter f(t) is a linear function or a convex function.

[0020] Preferably, the metabolic parameter-based feed medium development method adopts a linear function for f(t).

[0021] According to another aspect of the present application, a metabolic parameter-based feed medium development system is provided, which comprises a metabolic parameter fitting module and a feed medium development module.

[0022] The metabolic parameter fitting module is configured to fit the metabolic parameters of cell culture to a preset time t0by using a metabolic parameter fitting model for a selected basal medium formula and cell culture target parameters. The metabolic parameter is the consumption amount of each ingredient of the basal medium when the cell culture is performed according to the formula of the basal medium to a preset time, which is submitted to the feed medium development module;

[0023] The feed medium development module estimates the metabolic consumption amount of each medium ingredient at the feed medium addition time based on the metabolic parameters of cell culture to a preset time and the cell state at the feed medium addition time, and determines the concentration of each ingredient of the feed medium so that the ingredients of the feed medium used at this time meet the metabolic consumption amount of each medium ingredient at this time.

[0024] According to another aspect of the present application, a non-transitory computer readable storage medium is provided, which stores a computer program, and the computer program is executed by a processor to implement the metabolic parameter-based feed medium development method provided by the present application.

[0025] Overall, compared with the prior art, the above technical solutions conceived by the present application can achieve the following beneficial effects:

[0026] The method for developing a feed medium based on metabolic parameters provided by the present application adopts a metabolic parameter fitting model, estimates the consumption rate of each component of a basic medium for a cell culture target, and develops a feed medium, so that the components of the feed medium meet the metabolic consumption of cell culture.

[0027] The present application can reduce repetitive experiments, effectively improve analysis efficiency, and thus accelerate research and development speed and save time. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 is a process flow diagram of the method for developing a feed medium based on metabolic parameters provided by the present application;

[0029] Figure 2 is a scatter plot of each basic medium component and protein expression amount on the seventh day provided by an embodiment of the present application. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application. In addition, the technical features involved in each embodiment of the present application described below can be combined with each other as long as they do not conflict with each other.

[0031] The method for developing a feed medium based on metabolic parameters provided by the present application, as shown in Figure 1 includes the following steps:

[0032] (1) Metabolic parameter fitting: for the selected basic medium formula and cell culture target parameters, a metabolic parameter fitting model is used to fit the metabolic parameters of cell culture to a preset time t0 The cell culture target parameters are the target cell states required for cell culture, the cell states are cell viability, cell density and / or biochemical indicators, the biochemical indicators are protein expression amount, glucose, lactic acid, ammonia and / or glutamine content; the metabolic parameters are the consumption amounts of each component in the basic medium when the cell culture is carried out according to the formula of the basic medium to the preset time;

[0033] The metabolic parameter fitting model inputs the basal medium formula and the cell culture target parameter, and outputs the metabolic parameter. The applicable fitting model includes but is not limited to: linear regression model, support vector machine regression model, K nearest neighbor model, XGBoost, ridge regression, LightGBM, random forest, GBDT, or deep learning model. The deep learning model includes but is not limited to: fully connected neural network, convolutional neural network, recurrent neural network or attention model. Preferably, the linear regression model.

[0034] The metabolic parameter fitting model is obtained by the following method:

[0035] (1-1) Collect training formula:

[0036] The training formula is formed by searching within the addition range of each component, including but not limited to the following three methods: randomly generating formula, DOE experimental design formula, and mixed formula.

[0037] The randomly generated formula is that for each component in the basal medium formula, a value is randomly taken within its addition range to form a basal medium sample formula.

[0038] (1-2) Obtain metabolic parameter data:

[0039] Using the basal medium formula formed in step (1-1), the cell culture experiment is cultured to a preset time t0, the residual concentration of each component of the training formula obtained in step (1-1) is obtained, and the difference between the starting concentration and the residual concentration of each component of the culture medium is calculated as the consumption data of each component of the basal medium, i.e. metabolic parameter; It should be noted that the glucose level is maintained at the initial addition amount of the basal medium in the cell culture experiment.

[0040] (1-3) Organize training sample data set:

[0041] The training sample data set is organized by taking the addition amount of each component of the training formula obtained in step (1-1) and the original data or normalized value of the cell culture target parameter as the input matrix, and taking the metabolic parameter data obtained in step (1-2) as the output matrix.

[0042] (1-4) The metabolic parameter fitting model is obtained by training the fitting model with the training sample data set obtained in step (1-3).

[0043] In the preferred scheme, the cell culture target parameter is a biochemical indicator, and the metabolic parameter fitting model is a linear model. The biochemical indicator is the content of a certain substance, such as the expression amount of a specific protein or the number of cells. Since the conservation of mass is followed in biochemical experiments, the biochemical indicator and the metabolic parameter show a linear relationship, and a linear model is suitable for simulation.

[0044] The base medium formula can be recommended or experimentally determined by a base medium development system based on artificial intelligence.

[0045] (2) Feed medium development: according to the metabolic parameters of the cells cultured for a preset time obtained in step (1) and the cell state at the feed medium addition time, the metabolic consumption amounts of each medium component at the feed medium addition time are estimated; the concentrations of the components of the feed medium are determined so that the components of the feed medium used at this time meet the metabolic consumption amounts of each medium component at this time;

[0046] Preferably, the function f(t) is a linear function or a convex function, preferably a linear function. The metabolic consumption amounts of each medium component at the feed medium addition time t are estimated Where t=0, f(t)=[0, 0,...], i.e. when the culture time is 0, the consumption amount of each component of the medium is also 0; t=t0, i.e. when the culture is cultured for a preset time t0, the consumption amount of each component of the medium is the metabolic parameter f(t) is a linear function or a convex function, preferably a linear function.

[0047] In theory, the consumption amount of each component of the medium is adapted to the cell growth curve, and the consumption amount of each component of the medium within a certain time is a concave function curve that rises relatively quickly, so the metabolic consumption amount of each medium component at the feed medium addition time t is estimated by a linear function or a convex function The estimated metabolic consumption amount of each medium component Should exceed the actual metabolic consumption amount of each medium component, so that the feed medium developed accordingly can fully meet the needs of cell growth.

[0048] Estimating the metabolic consumption amount of each medium component at the feed medium addition time by a linear function is relatively simple compared to a convex function, and will not cause excessive enrichment of each component of the medium.

[0049] The following is an example:

[0050] The present application provides a medium formula development method based on the combination of metabolic parameters and artificial intelligence, as shown in Figure 1 The method comprises the following steps:

[0051] (1) Metabolic parameter fitting: for the selected base medium formula and the cell culture target parameters, a metabolic parameter fitting model is used to fit the metabolic parameters of the cells cultured for a preset time t0 The cell culture target parameter is a target cell state required for cell culture, the cell state is cell viability, cell density, and / or a biochemical index, and the biochemical index is protein expression amount, glucose, lactic acid, ammonia, and / or glutamine content; the metabolic parameter is the consumption amount of each component in the basal medium when the cell culture is performed according to the formula of the basal medium to a preset time.

[0052] The selected basal medium process is as follows:

[0053] The cell culture test is performed according to the training formula and the preset process:

[0054] The training formula is collected: according to the cell characteristics, a batch of formulas is screened from the formula library as a sample formula database of the to-be-developed formula.

[0055] The sample formula database has been established by the company, and the number of formulas is more than 1000.

[0056] The cell culture test is performed in a batch culture manner, the batch culture method is as follows: inoculation is performed at a cell density of 0.5x106 cells / mL, the culture volume is 10 mL, the culture container is a 50 mL mini bioreactor, the shaking speed is 180 rpm, the culture time is seven days, sampling is performed at the third day, the fifth day, and the seventh day during the culture process, the cell density is counted, and biochemical parameters such as glucose, lactic acid, ammonia, glutamine, and protein expression amount are detected, and glucose is supplemented to 4-5 g / L according to the glucose consumption. In order to obtain complete test data, sampling will be performed every day.

[0057] The target cell is cultured by using the basal medium sample formula, the cell state is detected by sampling at time points during the culture process, the cell state includes cell viability, cell density, and / or a biochemical index, and the biochemical index is protein expression amount, glucose, lactic acid, ammonia, and / or glutamine content; the cell viability can be fitted to obtain a cell viability curve of the basal medium sample formula with respect to the culture time, and the cell density is fitted to obtain a cell growth curve of the basal medium sample formula with respect to the culture time; the culture effect of the basal medium sample formula is the cell growth curve, the cell viability curve, or the cell density, the cell viability, and the biochemical index at a specific time point of the basal medium with respect to the culture time, which can also be the above multiple indexes or a comprehensive index of the above multiple indexes.

[0058] According to the culture effect of the above basal medium sample formula, a basal medium development system for the corresponding culture effect can be developed through sample training.

[0059] Determine the base medium formula: for the preset culture index of the preset culture time, use the base medium development system to predict the base medium composition and the index, recommend the base medium formula and select it, and obtain the selected base medium formula;

[0060] The training data of the metabolic parameter fitting model is obtained by the following method:

[0061] (1-1) Collect training formulas

[0062] In the training formulas collected in the selected base medium, 30-500 formulas are selected for constructing the metabolic parameter model, which are derived from the sample formula database. When the total amount of samples is more than 30, the components of metabolic parameters can be obtained in more than 5. Preferably, the selected scheme in this embodiment, the total amount of samples is more than 100, and the components of metabolic parameters can be obtained in more than 10.

[0063] (1-2) Obtain the metabolic component consumption data of the metabolomics medium: obtain part of the medium formula from the sample formula database for inoculation culture, perform experiments according to the development purpose, and at the same time, collect the metabolic conditions of each medium component on the seventh day by using mass spectrometer, liquid phase and other equipment, and establish a metabolic parameter database. For the components of the base medium sample formula, determine the nuclear mass ratio of each component, and according to the peak value under different nuclear mass ratios, obtain the metabolic parameters of the corresponding components. The mass spectrometer can obtain the residual content of part of the medium formula components; further, by using the component content of the medium formula design, the consumption content of each component of the cell can be obtained. For the consumption amount of different components and the culture effect, the visualization result is as shown in Figure 2 , wherein xi represents the consumption amount of different components, i.e. metabolic parameters, and yi represents the culture effect of the medium.

[0064] Figure 2 The scatter plot relationship between different component consumptions is shown.

[0065] (1-3) Organize the training sample data set: for sample data, data cleaning mainly uses the following methods: missing value processing, outlier processing, data set balancing, data noise processing, data deduplication, and data format standardization. The original data or normalized values of the training formula components obtained in step (1-1) and the cell culture target parameters are used as the input matrix, and the metabolic parameter data obtained in step (1-2) is used as the output matrix to organize the training sample data set; in this embodiment, the protein expression amount on the seventh day is used as the cell culture target parameter.

[0066] (1-4) Using the training sample data set obtained in step (1-3), the metabolic parameter fitting model is trained to obtain the metabolic parameter fitting model.

[0067] Metabolic parameter fitting model training:

[0068] The embodiment adopts python language, and linear regression models are respectively constructed for the consumption amount of different components of the selected basic medium and the protein expression amount on the seventh day. The component consumption amount is taken as the input value of the model, and the protein expression amount on the seventh day is taken as the output value. In the case of sufficient data amount, the model can accurately predict the protein expression amount under different component consumption amounts.

[0069] (2) Development of feed medium: according to the metabolic parameters of the cells cultured for a preset time obtained in step (1), and the cell state at the addition time of the feed medium, the metabolic consumption amount of each medium component at the addition time of the feed medium is estimated; and the components of the feed medium used at this time are made to meet the metabolic consumption amount of each medium component at this time;

[0070] The embodiment adopts a linear function Estimating the metabolic consumption amount of each medium component at the addition time t of the feed medium Wherein t=0, f(t)=[0, 0,...], that is, when the culture time is 0, the consumption amount of each component of the medium is also 0; t=t0, That is, the consumption amount of each component of the medium is the metabolic parameter when the culture time is preset to seven days t0. f(t) is a linear function or a convex function, preferably a linear function.

[0071] Specifically, for the metabolic parameter fitting model obtained by linear fitting in the embodiment, the relationship between the input value and the output value is analyzed. According to the linear regression model, the slope parameter of the model can be obtained. According to the slope parameter, it can be obtained that how much the input value increases, and at the same time, the corresponding output value also increases. Taking the yield of the target product as an example, according to the model constructed according to the consumption amount of different components, it is obtained that the unit yield of the target product corresponds to a specific amount of component consumption, as shown in Table 1.

[0072] Table 1: Different component consumption amounts mg / L corresponding to an increase of 100 mg / L of protein expression amount

[0073] x1 x2 x3 x4 x5 x6 y 38 63 33 20 15 51

[0074] Estimating the metabolic consumption amount of each medium component at the addition time of the feed medium specifically includes the following steps:

[0075] S1, estimating the metabolic consumption amount of each medium component of the medium per day:

[0076]

[0077] Wherein C Ttter is the target protein expression amount, such as a target protein expression amount of 6000 mg / L, and Days is the culture time, such as a culture time of 15 days, CComponent / Day To express the daily consumption of each component of the cell, which is also the optimal daily component addition amount; To express the specific amount of component consumption corresponding to the unit yield; and ΔC Component It can be expressed as the consumption of multiple different components, such as the presence of alternative components, and the like, and can express the common consumption of two or more associated components. For example, the daily protein expression amount change is 400 mg / (L*d), according to Table 1, for each increase of 100 units of y, 38 units of x1 are consumed, and the daily component consumption of x1 is 152 mg / (L*d).

[0078] S2, according to the process flow method of cell culture, determine the optimal component addition amount of the cell for multiple days. If the daily feed, then multiply 1 on the basis above, if every other day feed, multiply 2. Therefore, n-day feed, calculate the optimal component addition amount for n days, the formula is as follows:

[0079] C nDayComponent = n*C Component / Day

[0080] Wherein C nDayComponent is the optimal component addition amount for n days. For example, if every other day feed, and the daily component consumption of x1 is 152 mg / (L*d), then the optimal component addition amount for every other day is 304 mg / L.

[0081] S4, finally, for different process flows, the volume of the liquid preparation also affects the concentration. The base medium does not need to be adjusted for dilution factor, while the feed medium is related to the dilution factor. Taking the feed medium as an example, calculate the feed to be added, the concentration formula of every other day feed is as follows:

[0082]

[0083] Wherein C FeedComponent represents the component content of the feed medium, V Inittal , V Feed respectively represent the initial medium volume and the feed medium volume. For example, taking 5% of the original solution volume as the feed volume, the optimal component addition amount for every other day is 304 mg / L, and the optimal component concentration of the feed is 6383 mg / L.

[0084] Through the above four steps, the component content of the medium formula based on the metabolic parameter data can be obtained.

[0085] x1 x2 x3 x4 x5 x6 6383 10584 5543 3360 2520 8568

[0086] For the component content of the medium that cannot be detected, refer to the high-quality medium formula of the sample formula database to define the appropriate component content. Thus, a complete medium formula system is developed.

[0087] It is to be understood that the above description is intended to be illustrative and not restrictive. Many other embodiments will be apparent to those of skill in the art upon reading and understanding the above description. The scope of the application should, therefore, be determined with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled.

Claims

1. A method for developing a feed medium based on metabolic parameters, characterized by, The method comprises the following steps: (1) Metabolic parameter fitting: for the selected basal medium formula and the cell culture target parameter, a metabolic parameter fitting model is used to fit the cell culture to a preset time metabolic parameters ; the metabolic parameters are the consumptions of the components in the basal medium when the cell culture is performed according to the formula of the basal medium to a preset time; The cell culture target parameter is a target cell state required by cell culture; the cell state is cell viability, cell density, and / or a biochemical index, and the biochemical index is protein expression amount, glucose, lactic acid, ammonia, and / or glutamine content; The metabolic parameter fitting model inputs a basal medium formula and the cell culture target parameter; The output of the metabolic parameter fitting model is a metabolic parameter; (2) Feed medium development: according to the metabolic parameter of cell culture to a preset time and the cell state at the feed medium addition time obtained in step (1), metabolic consumption amounts of each medium component at the feed medium addition time are estimated; the concentrations of each component of the feed medium are determined, so that the components of the feed medium used at this time meet the metabolic consumption amounts of each medium component at this time.

2. The method for developing a feed medium based on metabolic parameters according to claim 1, wherein, The metabolic parameter fitting model comprises a linear regression model, a support vector machine regression model, a K-nearest neighbor model, XGBoost, ridge regression, LightGBM, random forest, GBDT, or a deep learning model; the deep learning model comprises a fully connected neural network, a convolutional neural network, a recurrent neural network, or an attention model.

3. The method for developing a feed medium based on metabolic parameters according to claim 1 or 2, characterized in that, The cell culture target parameter is a biochemical index, and the metabolic parameter fitting model is a linear model.

4. The method for developing a feed medium based on metabolic parameters according to claim 1, wherein, The metabolic parameter fitting model is obtained according to the following method: (1-1) Collect training formulas; (1-2) Obtain metabolic parameter data: Using the basal medium formulation formed in step (1-1), culture to a predetermined time Carrying out cell culture experiments, obtaining the residual concentrations of each component of the training formulation obtained in step (1-1), and calculating the difference between the starting concentration and the residual concentration of each component of the culture medium as the consumption data of each component of the basal medium, i.e. metabolic parameters; (1-3) Organize a training sample data set: The training sample data set is organized by taking the addition amounts of each component of the training formula obtained in step (1-1) and the original data or normalized values of the cell culture target parameter as an input matrix and taking the metabolic parameter data obtained in step (1-2) as an output matrix; (1-4) The training sample data set obtained in step (1-3) is used to train a fitting model to obtain the metabolic parameter fitting model.

5. The method for developing a fed-batch medium based on metabolic parameters according to claim 1, wherein, Adopting function Estimating feeding medium addition time The metabolic consumption of each medium component Wherein , That is, when the culture time is 0, the consumption of each component of the medium is also 0; , That is, when the culture time is preset time , the consumption of each component of the medium is metabolic parameter ; It is a linear function or a convex function.

6. The method for developing a feed medium based on metabolic parameters according to claim 5, wherein, The is a linear function.

7. A metabolic parameter based feed medium development system, characterized in that, The method comprises: a metabolic parameter fitting module and a feed medium development module; The metabolic parameter fitting module is configured to fit cell culture to a preset time using a metabolic parameter fitting model for a selected basal medium formula and cell culture target parameters metabolic parameters ; the metabolic parameters are consumptions of components in the basal medium when the cell culture is performed to the preset time according to the formula of the basal medium, which are submitted to the feed medium development module; The cell culture target parameter is a target cell state required by cell culture; the cell state is cell viability, cell density, and / or a biochemical index, and the biochemical index is protein expression amount, glucose, lactic acid, ammonia, and / or glutamine content; The metabolic parameter fitting model inputs a basal medium formula and the cell culture target parameter; The output of the metabolic parameter fitting model is a metabolic parameter; The feed medium development module estimates metabolic consumption amounts of each medium component at the feed medium addition time according to the metabolic parameter of cell culture to a preset time and the cell state at the feed medium addition time; the concentrations of each component of the feed medium are determined, so that the components of the feed medium used at this time meet the metabolic consumption amounts of each medium component at this time.

8. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by a processor, implements the feed medium development method based on metabolic parameters according to any one of claims 1 to 6.

Images