Method for estimating metabolic flux distribution, analysis device, and program

Abstract

Provided is a method for estimating the metabolic flux distribution of cells, the method comprising: a step for acquiring discharge data including time-series data of a substance discharged to the outside of the cells with respect to a first strain cell and a second strain cell that is a reference for comparison; a step for calculating a discharge flux of the substance to the outside of the cells on the basis of the discharge data; a step for estimating a plurality of primary flux patterns on the basis of the discharge flux; and a step for clustering the plurality of primary flux patterns on the basis of the mutual similarity of the plurality of primary flux patterns and then outputting the clustered result.

Description

Method for estimating metabolic flux distribution, analysis device, and program

[0001] The present invention relates to a method, an analysis device, and a program for estimating metabolic flux distribution, and more particularly to a method, an analysis device, and a program for estimating metabolic flux distribution of a cell.

[0002] There is a field of synthetic biology in which the metabolism of cells such as microorganisms is modified using genetic engineering techniques to produce useful substances for use in fuels, functional foods, pharmaceuticals, etc.

[0003] Non-Patent Documents 1 and 2 disclose a method for designing cellular metabolism using flux balance analysis (FBA) in the field of synthetic biology. FBA is a metabolic flux analysis method that constructs a metabolic model in which metabolic reactions are expressed in a linear form and determines the metabolic flux distribution that maximizes the cell growth rate using linear programming.

[0004] Matsuda, F., Yoshikawa, K., Shimizu, H., "Intracellular Metabolic Design and Engineering Applications by Metabolic Simulation," Bioengineering, Vol. 92, 593-597 (2014); Toya, Y., Metabolic Design Technology Using Computer Simulation, Chemistry and Biology, Vol. 55, 83-85 (2017); A. K. Ramirez et al., "Integrating Extracellular Flux Measurements and Genome-Scale Modeling Reveals Differences between Brown and White Adipocytes," Cell Reports, Vol. 21, 3040-3048 (2017). Nobuyuki Okahashi, Shuichi Kawana, Fumio Matsuda, Hiroshi Shimizu, "Introduction to 13C metabolic flux analysis using GC-MS", Technical Report (C146-0355). Hiroshi Shimizu, Fumio Matsuda, Yoshihiro Toya, "Application of metabolic flux analysis using metabolic design and 13C isotope labeling to material production", Kagaku to Seibutsu 53 (7): 455-461 (2015)

[0005] In metabolic flux analysis methods such as FBA, attempts have been made to determine metabolic flux distributions by imposing parameter constraints based on the actual measured values ​​of several metabolites. Non-Patent Document 3 discloses a method for performing FBA using the actual measured values ​​of four added inhibitors and two metabolites excreted outside the cell. This allows for the estimation of metabolic flux distributions for the actually measured fluxes of several metabolites that closely match the metabolic reactions actually occurring in the cell.

[0006] However, the flux solution space in metabolic models is enormous, and metabolic flux analyses such as FBA generally extract only one solution from the vast solution space. Therefore, it is possible that conventional methods may not be able to accurately estimate the fluxes of substances other than the few metabolites mentioned above.

[0007] The present disclosure has been made in view of the above circumstances, and its purpose is to provide a method for correctly estimating metabolic flux in the analysis of metabolic flux distribution in cells.

[0008] A first aspect of the present disclosure is a method for estimating a metabolic flux distribution of a cell, comprising the steps of: acquiring excretion data including time series data of substances excreted outside the cell for each of a first cell line and a second cell line for comparison; calculating the excretion flux of the substance outside the cell based on the excretion data; estimating a plurality of first-order flux patterns based on the excretion flux; and clustering and outputting the plurality of first-order flux patterns based on the similarity between the plurality of first-order flux patterns.

[0009] A second aspect of the present disclosure is a method for estimating a metabolic flux distribution of a cell, comprising the steps of: acquiring excretion data for a predetermined cell line, the excretion data including time series data of substances excreted outside the cell; calculating the excretion flux of the substance outside the cell based on the excretion data; estimating a plurality of first-order flux patterns based on the excretion flux; and clustering and outputting the plurality of first-order flux patterns based on the similarity between the plurality of first-order flux patterns.

[0010] A third aspect of the present disclosure is an analysis device for estimating a metabolic flux distribution of a cell. The analysis device includes a memory and a processor. The memory stores excretion data for each of a first cell line and a second cell line as a comparison target, the excretion data including time-series change data of substances excreted outside the cell. The processor calculates the excretion flux of the substance to the cell based on the excretion data, estimates multiple first-order flux patterns based on the excretion flux, and clusters and outputs the multiple first-order flux patterns based on the similarity between the multiple first-order flux patterns.

[0011] According to the method for estimating metabolic flux distribution according to the present disclosure, it is possible to provide a method for correctly estimating metabolic flux in the analysis of metabolic flux distribution in cells.

[0012] Fig. 1 is a diagram showing a configuration of an analysis system according to an embodiment. Fig. 2 is a diagram for explaining a metabolic map. Fig. 3 is a diagram for explaining metabolic flux distribution and FBA. Fig. 4 is a diagram for explaining a method for estimating a flux pattern according to an embodiment. Fig. 5 is a flowchart showing a process for estimating a flux pattern according to an embodiment. Fig. 6 is a diagram for explaining a method for calculating proliferation flux. Fig. 7 is a flowchart showing a process for estimating a primary flux pattern.

[0013] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described in detail below with reference to the accompanying drawings. In the following description, the same or corresponding parts in the drawings are denoted by the same reference numerals, and their description will not be repeated in principle.

[0014] 1. Configuration of Analysis System] Fig. 1 is a diagram showing the configuration of an analysis system according to an embodiment. Referring to Fig. 1, analysis system 100 includes an analysis device 1, a measurement device 2, and a culture device 3.

[0015] The culture device 3 is a device for culturing cells. In one embodiment, the culture device 3 cultures a first cell line, which is a first cell line, and a second cell line, which is a second cell line and is a comparison target for the first cell line. In one embodiment, the first cell line and the second cell line are cell lines that differ in the amount of production of a specific type of metabolite (hereinafter referred to as a "specific metabolite"). Furthermore, the first cell line and the second cell line are cells that belong to two types of strains of the same lineage. Furthermore, the specific metabolite is a specific useful metabolite that is intended to be obtained by culture. In this specification, such a specific useful metabolite is referred to as a "target product."

[0016] The measuring device 2 measures the amount of metabolites excreted outside the cells. In this specification, metabolites excreted outside the cells are also referred to as "extracellular metabolites."

[0017] The measuring device 2 includes, for example, a cell measuring unit 21 and an extracellular metabolite measuring unit 22. The cell measuring unit 21 measures the amount of cells (hereinafter also referred to as "cell mass"). The cell measuring unit 21 may directly measure the cell mass or may calculate the cell mass by measuring the concentration of a predetermined substance correlated with the cell mass. As an example, the cell measuring unit 21 includes a turbidity meter, in which case the cell mass can be measured based on turbidity. Another example of the cell measuring unit 21 includes an instrument for measuring the dry cell weight of cells (e.g., a drying dish and a weighing scale). In this case, the user measures the dry cell weight of the cells using the instrument and then measures the cell mass based on the measured dry cell weight. Yet another example of the cell measuring unit 21 includes a predetermined instrument and chemicals such as RIPA (Radio-Immunoprecipitation Assay) buffer and bicinchoninic acid (BCA). In this case, the user uses these chemicals to extract and quantify proteins, thereby quantifying the cell mass. In addition, when the cell mass is correlated with the amount of a predetermined type of extracellular metabolite, the measurement of the cell mass may be replaced by the measurement of the amount of the predetermined type of extracellular metabolite.Relatedly, when the cell mass is correlated with the consumption amount of a predetermined type of substrate, the measurement of the cell mass may be replaced by the measurement of the consumption amount.The predetermined type of substrate is, for example, oxygen.

[0018] In one embodiment, the cell measuring unit 21 also records the time at which the cell mass is measured. In this case, for example, the cell measuring unit 21 transmits time-series data of the cell mass, including the time and the cell mass corresponding to the time, to the analysis device 1. The time-series data includes time-series changes. The time-series data of the cell mass is hereinafter also referred to as "cell data." However, the method of acquiring the cell data in the analysis system 100 is not limited to the above example. For example, the cell measuring unit 21 may immediately transmit the measured value of the cell mass to the analysis device 1, and the analysis device 1 may calculate the time at which the cell mass was measured based on the time at which the measured value of the cell mass was received, and create data on the cells during culture.

[0019] The extracellular metabolite measuring unit 22 is a unit for measuring the amount of extracellular metabolites (hereinafter also referred to as "extracellular metabolite amount"). The extracellular metabolite measuring unit 22 includes a device for measuring the amount of extracellular metabolites. The device for measuring the amount of extracellular metabolites includes, for example, an analytical device such as a GC (Gas Chromatograph), an LC (Liquid Chromatograph), an MS (Mass Spectrometry), an LC-MS (Liquid Chromatograph-Mass Spectrometry), or an HPLC (High Performance Liquid Chromatography). For example, a user extracts a portion of the culture solution in the culture device 3 and uses the extracted portion of the culture solution as a sample to analyze them using the analytical device included in the extracellular metabolite measuring unit 22.

[0020] In one embodiment, the extracellular metabolite measuring unit 22 also records the time at which the cells are analyzed. In this case, for example, the extracellular metabolite measuring unit 22 transmits time-series data of extracellular metabolite amounts, including the time and the extracellular metabolite amounts corresponding to the time, to the analysis device 1. The time-series data of extracellular metabolite amounts will hereinafter also be referred to as "extracellular metabolite data." However, the method of acquiring extracellular metabolite data in the analysis system 100 is not limited to the above example. For example, the extracellular metabolite measuring unit 22 may immediately transmit the analyzed values ​​of the extracellular metabolites to the analysis device 1, and the analysis device 1 may calculate the time at which the extracellular metabolite amounts were measured based on the time at which the measured values ​​of the extracellular metabolite amounts were received, and create the extracellular metabolite data.

[0021] The analysis device 1 acquires cell data and extracellular metabolite data, and estimates a "flux pattern," which is a pattern of cellular metabolic flux distribution, based on these data. The method for estimating the flux pattern will be described later.

[0022] The analysis device 1 includes a controller 19, a display 15, and an operation unit 14. The display 15 and operation unit 14 are connected to the controller 19. The operation unit 14 is typically composed of a touch panel, a keyboard, a mouse, etc. The operation unit 14 accepts user operation input to the processor 10. The display 15 is composed of, for example, a liquid crystal panel capable of displaying images. The display 15 displays images related to the acceptance of the user operation input and displays the results of processing by the processor 10.

[0023] The controller 19 has, as its main components, a processor 10, a memory 11, a communication unit 12, and an input / output unit 13. These units are connected to each other via a bus so as to be able to communicate with each other.

[0024] The input / output unit 13 is an interface for exchanging various types of data between the processor 10 and external devices connected to the input / output unit 13. The external devices include the measurement device 2, an operation unit 14, and a display 15. In one embodiment, the input / output unit 13 receives the extracellular metabolite data and the cell data from the measurement device 2. However, the method for acquiring the extracellular metabolite data and the cell data in the analysis device 1 is not limited to this, and the data may be acquired via the communication unit 12 or via a storage medium such as a USB memory.

[0025] The communication unit 12 is a communication interface for exchanging various data with an external device, and is realized by an adapter, a connector, etc. The communication method may be a wireless communication method using a wireless LAN (Local Area Network) or the like, or a wired communication method using a USB (Universal Serial Bus) or the like.

[0026] The memory 11 stores the extracellular metabolite data and cell data acquired by the input / output unit 13. The memory 11 is realized by a storage device such as a read-only memory (ROM), a random access memory (RAM), and a hard disk drive (HDD). The ROM can store programs executed by the processor 10. The RAM can temporarily store data used during program execution by the processor 10 and can function as a temporary data memory used as a working area. The HDD is a non-volatile storage device. A semiconductor storage device such as a flash memory may be used in addition to or instead of the HDD. The programs and / or data may be stored in an external storage device accessible by the processor 10.

[0027] The processor 10 estimates a flux pattern, which is a pattern of metabolic flux distribution in a cell, based on the extracellular metabolite data and cell data stored in the memory 11. The processor 10 is typically an arithmetic processing unit such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit). The processor 10 controls the operation of the analysis device 1 by reading and executing programs stored in the memory 11. The programs include a program that, when executed by a computer, causes the computer to implement a flux pattern estimation method according to an embodiment.

[0028] [2. Conventional Methods for Estimating Metabolic Flux Distribution and Their Problems] Living organisms' cells take in substrates from the outside, convert them into various metabolites, and excrete some of them to the outside. Taking advantage of this phenomenon, technologies have been researched and developed in the field of synthetic biology to modify cells, for example by genetically manipulating them, so that they appropriately produce specific, useful metabolites of interest.

[0029] In the field of synthetic biology, changes in cellular metabolism due to changes in culture conditions, genetic mutations due to subcultures, etc., can cause changes in productivity of target products. To investigate the causes of changes in productivity, it is important to understand metabolic flow rate. Metabolic flow rate is called "metabolic flux" or simply "flux."

[0030] (2-1. Overview of FBA) To understand fluxes, FBA may be used, as explained in Non-Patent Documents 1 and 2. FBA is a method for constructing a metabolic model that linearly represents metabolic reactions and estimating metabolic flux distributions using linear programming. FBA will be described in more detail below with reference to FIGS. 2 and 3.

[0031] FIG. 2 is a diagram for explaining the "metabolic map" that is the premise of FBA. FIG. 2 shows a schematic map (metabolic map) of metabolic reactions that can occur in a cell (C). Circles (M) indicate each metabolite, and arrows indicate the conversion of the metabolite. The metabolic map is created, for example, based on the genomic information of the cell.

[0032] First, cells take up a substrate. Then, they convert the taken-up substrate into another metabolite. After repeated conversion of metabolites within the cell, some metabolites are eventually excreted outside the cell. In the field of synthetic biology, some of the metabolites excreted outside the cell are often used as target products. Metabolites other than the target product excreted outside the cell are sometimes called by-products. In addition, some of the metabolites are used to construct the next generation of cells.

[0033] Although such metabolic maps can be created based on genome information as described above, it is difficult to predict which metabolic reactions are actually occurring and to what extent. FBA is a method for predicting the extent of such metabolic reactions that are actually occurring.

[0034] Figure 3 illustrates metabolic flux distribution and FBA. Flux is expressed as the amount of change in metabolites. In the example of Figure 3(A), the amount of change v1 from the substrate to the metabolite indicated by a circle M1, or the amounts of change v2 and v3 from the metabolite to other metabolites are illustrated.

[0035] When performing FBA, two assumptions are set as conditions. The first condition for FBA is that the metabolism of all metabolites in the cell is in a steady state. This means that the amount of all metabolites does not change over time. Therefore, in the mass balance equation for metabolite M1 shown in Figure 3(A), the change, v1 - v2 - v3, is 0. In this way, a mass balance equation with a solution of 0 is established for all metabolites.

[0036] The mass balance equations for all metabolites are combined into a single determinant. A model that comprehensively includes cellular fluxes, such as this determinant, is referred to as a metabolic model, a stoichiometric model, or the like by those skilled in the art. In this specification, however, it is referred to as a "metabolic model." In FBA, a metabolic model based on this first condition is used to uniquely determine the flux distribution that satisfies the second condition, i.e., the cell growth rate is maximized ( FIG. 3B ).

[0037] In this way, by imposing parameter constraints on the flux distribution created by FBA, such as environmental conditions, the presence or absence of disruption of a specific gene, nutritional requirements, etc., it is possible to estimate changes in metabolic flux distribution. Parameter constraints include, for example, upper and lower limits for each flux. Examples have been reported in which changes in the flux distribution estimated in this way were verified by applying changes in environmental conditions, etc., to actual microorganisms.

[0038] However, conversely, if parameter constraints corresponding to actual environmental conditions are not applied to the flux distribution created by FBA, incorrect changes in the flux distribution will be estimated. Therefore, it is important to apply parameter constraints corresponding to actual environmental conditions in FBA.

[0039] From this perspective, research has been reported that estimates flux distributions that fit actual measured values ​​by performing FBA while providing the actually measured amounts of extracellular metabolites as parameter constraints for a metabolic model.

[0040] (2-2. FBA using measured values ​​and its problems) The unit of flux is expressed as mmol / (h gDCW), where the unit is cell mass (gram dry cell weight), unit time (h), and millimolar mass (mmol) of metabolic component. Therefore, to calculate flux from actual measured values, it is necessary to obtain time-series data of metabolite concentration (mol / mL) and cell density (gDCW / mL). Analytical devices such as HPLC and MS are used to measure metabolite concentration. Cell density can be measured by measuring the culture medium with a turbidity meter, or by extracting and quantifying protein using RIPA buffer and BCA.

[0041] Non-Patent Document 3 discloses an example of FBA performed using actual measurements of four inhibitors and two extracellularly excreted metabolites. In the example in Non-Patent Document 3, FBA was performed using the directly measured flux values ​​measured using a Seahorse XF Analyzer and the indirectly measured flux values ​​calculated from the measured values ​​as flux constraints in the metabolic model. As a result, in the estimated flux distribution, the estimated values ​​of GABA (Gamma-Aminobutyric Acid), ABAT (4-Aminobutyrate Aminotransferase), glucose, glutamate, and palmitate matched the actual measured values. However, the flux solution space in the metabolic model is enormous, and FBA extracts only one solution from that, so it is possible that fluxes related to substances other than the five substances evaluated above may not be accurately estimated.

[0042] (2-3. Problem of difficulty in identifying the metabolic pathway that is the cause from changes in measured values) Furthermore, in FBA, it is possible to perform forward calculations to change the productivity or growth rate of a target substance by imposing constraints on the flux of one of the metabolic pathways, but it has been difficult to identify the metabolic pathway that is the cause from the measured value of a change in productivity of a target substance.

[0043] Because a large number of metabolic flux distribution patterns ("flux patterns") that match the measured values ​​of productivity changes are calculated, there is also the problem that the work required for users to check the flux patterns becomes enormous.

[0044] (2-4. 13C metabolic flux analysis method and its problems) Flux analysis methods using stable isotopes are also known as a method for estimating flux in the field of synthetic biology. Non-Patent Documents 4 and 5 disclose a method for estimating the flux distribution of intracellular metabolic pathways using 13C metabolic flux analysis method using GC-MS. 13C is carbon-13, a stable isotope of naturally occurring carbon-12 (12C).

[0045] In the 13C metabolic flux analysis method, cells are cultured in a medium containing 13C glucose, a carbon source labeled with the stable carbon isotope 13C, and the flux distribution can be estimated by measuring the 13C-labeled ratio in metabolites based on 13C glucose taken up by the cells.

[0046] However, the 13C metabolic flux analysis method has the problem that 13C glucose is expensive. Furthermore, since glycerol, carbon dioxide, etc. cannot be labeled with 13C, it cannot be applied to hosts that use these substances as carbon sources. Furthermore, flux analysis methods using stable isotopes, including 13C metabolic flux analysis, require tuning of experimental conditions even with slight differences in the experimental system, which presents a high technical hurdle. Specifically, flux analysis methods using stable isotopes require the experimental system to be stopped at a desired state, which is technically very difficult.

[0047] As described above, in the field of conventional synthetic biology, several methods for estimating flux patterns have been known, but each has its own problems.

[0048] 3. Method for Estimating Flux Pattern According to Embodiment In view of the above circumstances, the method for estimating a flux pattern according to this embodiment includes the following four processes, namely, first to fourth processes.

[0049] 4 is a diagram illustrating a method for estimating a flux pattern according to an embodiment. A feature of the first process of the method for estimating a flux pattern according to this embodiment is that time-series data of many types of extracellular metabolites is acquired, preferably time-series data of extracellular metabolome data. This is because, in order to estimate a flux pattern that is highly consistent with the metabolic reactions occurring in actual cells, it is preferable to measure as many types of extracellular metabolites as possible, and ideally, it is preferable to comprehensively measure all types of measurable extracellular metabolites.

[0050] In the first process, time-series data on cells and extracellular metabolites are first obtained. From the perspective of material balance, cells and extracellular metabolites produced by cell proliferation are both forms of substances excreted outside the cells. Therefore, in this specification, cells and extracellular metabolites are collectively referred to as "excretes," and time-series data on cell mass and time-series data on extracellular metabolite mass are collectively referred to as "excretion data."

[0051] Next, the growth flux, which is the flux of cell growth rate, is calculated based on the cell data, and the extracellular flux, which is the change in growth flux, is calculated based on the extracellular metabolite data. In this specification, growth flux and extracellular flux are collectively referred to as "excretion flux." In addition, in contrast to extracellular flux, the flux of intracellular metabolic reactions is referred to as "intracellular flux."

[0052] The many types of extracellular metabolites preferably include amino acids, vitamins, organic acids, metabolites involved in amino acid metabolism (organic acids, etc.), metabolites involved in nucleic acid metabolism, metabolites involved in the TCA cycle (Tricarboxylic Acid Cycle), etc. The many types of extracellular metabolites are more preferably metabolites that constitute the totality of extracellular metabolites. The totality of extracellular metabolites is called the "extracellular metabolome," and the measured values ​​thereof are called "extracellular metabolome data."

[0053] The many types of extracellular metabolites include, for example, several tens or more types of substances, and preferably, one hundred or more types of substances.

[0054] In the second process, an optimization problem and flux variability analysis (FVA) are used to estimate multiple flux patterns based on the measured values. Specific calculation methods will be described later. The flux patterns estimated by the optimization problem and FVA are also referred to as "primary flux patterns" in this specification.

[0055] In the third and fourth processes, post-processing (clustering and ranking) is performed to present the most plausible flux pattern among the estimated primary flux patterns to the user. In other words, the most plausible flux pattern is one that is considered to be highly likely to occur in real cells.

[0056] In the third process, the estimated primary flux patterns are clustered into groups of similar primary flux patterns, and then a representative flux pattern is generated from the primary flux patterns in each cluster.

[0057] In the fourth process, the representative flux patterns are ranked in order of plausibility. In other words, the order of plausibility is the order in which the patterns are most likely to be performed in real cells. The method for ranking the representative flux patterns will be described later.

[0058] In the first process, the flux pattern estimation method according to this embodiment solves an optimization problem based on measured values ​​of cell proliferation rate and measured values ​​of many types of extracellular metabolites. Therefore, compared to FBA, which is constrained by the theoretical maximum cell proliferation rate, it is possible to obtain a flux pattern that is more consistent with the actual measured values. Furthermore, compared to FBA, which incorporates only measured values ​​of a small number (e.g., five types) of extracellular metabolites, it is possible to narrow down the flux space more. Furthermore, in the second process, instead of being limited to a single flux pattern as in FBA, it is possible to estimate multiple plausible flux patterns. Therefore, it is considered that the solution is more likely to include a flux pattern that reflects the metabolic reactions in actual cells than conventional metabolic flux analysis.

[0059] Furthermore, in the third and fourth processes, the user can rank and check representative patterns among a plurality of flux patterns, thereby not placing an excessive burden on the user.

[0060] Furthermore, according to the flux pattern estimation method of this embodiment, it is easy to estimate the metabolic pathway that is the cause of the change in productivity. Next, FIG. 5 shows the flux pattern estimation method of this embodiment.

[0061] [4. Flux Pattern Estimation Process According to the Embodiment] Fig. 5 is a flowchart showing a flux pattern estimation process according to the embodiment. The steps shown in Fig. 5 are executed by the processor 10 of the analysis device 1. Note that "S" is used as an abbreviation for "STEP" in the figure. S11 to S12 in Fig. 5 correspond to the first process in Fig. 4. S21 in Fig. 5 corresponds to the second process in Fig. 4. S31 in Fig. 5 corresponds to the third process in Fig. 4. S41 in Fig. 5 corresponds to the fourth process in Fig. 4.

[0062] Referring to FIG. 5, in S11, the processor 10 acquires excretion data including time-series change data of substances excreted outside the cells for each of the first cell line and the second cell line.

[0063] As described above, the second cell line has a different production amount of a specific metabolite from the first cell line. In one embodiment, the first cell line and the second cell line are cells belonging to two different strains of the same lineage.

[0064] The first and second cell lines may be prokaryotic or eukaryotic. The first and second cell lines are preferably derived from organisms for which a genome-based metabolic map has been constructed, such as Escherichia coli, Bacillus subtilis, Mycobacterium tuberculosis, Saccharomyces cerevisiae, green algae, nematodes, Arabidopsis thaliana, or humans. Such genome-based metabolic maps are also referred to as "genome-scale metabolic models" by those skilled in the art. However, even for cells derived from organisms for which no existing metabolic map exists, constructing a metabolic map makes it possible to use the metabolic flux pattern estimation method according to this embodiment.

[0065] The specific metabolite is typically a metabolite that is excreted outside the cell, and in this case, the "production amount of the specific metabolite" corresponds to the "excretion amount of the specific metabolite outside the cell." The specific metabolite is typically a target substance that is a predetermined useful metabolite.

[0066] The excretion data includes cell data, which is time series data of the amount of cells, and extracellular metabolite data, which is time series data of the amount of extracellular metabolites, which are metabolites excreted outside the cells. In one embodiment, the cell data is time series data of cell density (gDCW / mL), and the extracellular metabolite data is time series data of extracellular metabolite concentration (mol / mL).

[0067] In one embodiment, the time series data of the amount of the target substance is obtained from measurements by GC and / or MS. In one embodiment, the time series data of the amount of by-products (e.g., amino acids, vitamins, organic acids, metabolites involved in amino acid metabolism (e.g., organic acids), metabolites involved in nucleic acid metabolism, and metabolites involved in the TCA cycle) is obtained from measurements by an LC-MS cell culture profiling method (manufactured by Shimadzu Corporation).

[0068] In S12, the processor 10 calculates the excretion flux of the substance out of the cell based on the excretion data.

[0069] In one embodiment, the processor 10 calculates extracellular metabolite flux, which is the amount of change in extracellular metabolite per cell per time, from the extracellular metabolite data. Also, it calculates proliferation flux, which is the amount of change in extracellular metabolite per cell per time, from the cell data. The excretion flux includes the extracellular metabolite flux and proliferation flux.

[0070] Fig. 6 is a diagram illustrating a method for calculating the growth flux. Fig. 6 is a graph showing the time series change in cell mass, with the horizontal axis representing the culture time and the vertical axis representing the number of bacteria (log scale). In one embodiment, the growth flux is the value obtained by dividing the number of bacteria grown in a given period by the product of the number of bacteria in that given period and the number of bacteria at the start of the given period.

[0071] 5, in S21, the processor 10 estimates a plurality of primary flux patterns based on the exhaust flux. The process of estimating the primary flux patterns will be described in detail with reference to FIG.

[0072] In S31, the processor 10 estimates a representative flux pattern by clustering the plurality of primary flux patterns using information on the mutual similarity of the plurality of primary flux patterns. In this specification, the degree of similarity between the flux patterns or fluxes is referred to as "similarity."

[0073] Preferably, the processor 10 clusters multiple primary flux patterns based on the mutual similarity of the efflux fluxes. The reason for this is explained below. Even if the efflux fluxes are the same in primary flux patterns, the intracellular fluxes may differ depending on the degree of freedom of the intracellular fluxes. In this case, the difference in intracellular flux does not affect the efflux flux and is therefore not important as a factor in changing the productivity of the target product, so it is considered that it is not necessary to incorporate it into the clustering. This reduces the possibility that factors that are not important as a factor in changing the productivity of the target product will affect the clustering.

[0074] In one embodiment, the processor 10 performs clustering using hierarchical clustering. For example, the processor 10 calculates the Euclidean distance between two first-order fluxes, and repeats the process of clustering the pairs with the shortest distance (highest similarity) to create a tree diagram, thereby performing hierarchical clustering. For example, the distance calculation method uses Ward's method (minimum variance method).

[0075] The processor 10 estimates a representative flux pattern based on the primary flux patterns included in each cluster. The representative flux pattern may be selected from the primary flux patterns in the cluster, or may be created by integrating at least some of the primary flux patterns in the cluster.

[0076] In S41, the processor 10 ranks the representative flux patterns. The rank is a rank of plausibility, more specifically, a rank that is considered to be highly likely to be similar to a flux pattern actually occurring in a cell. In other words, the representative flux pattern ranked first is the most plausible flux pattern.

[0077] The ranking can be said to be a priority order for the user when considering representative flux patterns. For example, it is considered most efficient for the user to give top priority to the representative flux pattern ranked first in the ranking for verification experiments.

[0078] Preferably, the processor 10 ranks the representative flux patterns based on the consistency between the measured values ​​of efflux flux and the values ​​estimated by FBA. In one embodiment, the processor 10 ranks the representative flux patterns based on the error between the measured values ​​of growth flux and the values ​​estimated by FBA, and the error between the measured values ​​of extracellular metabolite flux and the values ​​estimated by FBA. The error is, for example, the root mean squared percentage error (RMSPE) calculated by the following formula:

[0079]

[0080] In S51, the processor 10 detects differences by comparing a representative flux pattern of the first cell line with a representative flux pattern of the second cell line. In one embodiment, the processor 10 first detects similar flux pattern pairs between the first cell line and the second cell line by comparing the representative flux pattern of the first cell line with the representative flux pattern of the second cell line. Next, the processor 10 compares the flux pattern pairs to detect differential fluxes that differ between the flux pattern pairs. The differential fluxes are fluxes that are estimated to be the cause of changes in the production yield of the target product in the second cell line. As described above, the user can estimate the fluxes that are the cause of changes in the production yield of the target product in the second cell line.

[0081] In S61, the processor 10 clusters and outputs the multiple primary flux patterns, and then ends the processing. In S61, the processor 10 may output at least one of a representative flux pattern, a ranked representative flux pattern, a flux pattern pair, and a difference flux. In one embodiment, the processor 10 outputs these estimation results by displaying them on the display 15. However, the output form of the estimation results is not limited to this, and the estimation results may be transmitted to an external device of the analysis system 100 via the communication unit 12 and output by the external device. The external device may be, for example, a printer or a computer with a display. This allows the user to easily understand the estimation results. Furthermore, the user can plan a verification experiment for the estimation results based on the estimation results.

[0082] 7 is a flowchart showing the process of estimating the primary flux pattern, which corresponds to the subroutine of S21 in FIG. 5 and is executed after S12 and before S51 in FIG.

[0083] The steps shown in Fig. 7 are executed by the processor 10 of the analysis device 1. Referring to Fig. 7, in S211, the processor 10 solves an optimization problem in which the intracellular flux is used as an explanatory variable, and the error between the measured and estimated values ​​of the cell proliferation rate and the error between the measured and estimated values ​​of the excretion amount of a specific metabolite are used as objective functions. The flux pattern estimated in this manner is referred to as an "optimized flux pattern" in this specification. In other words, the optimized flux pattern is a flux pattern optimized to explain the cell proliferation rate and the measured values ​​of a specific metabolite.

[0084] When solving an optimization problem, two conditions are set. The first condition, like the first condition in FBA, is that the mass balance equation must be 0 for all intracellular metabolites. The second condition, unlike the second condition in FBA, is that multiple flux patterns that minimize the error between the estimated and experimental values ​​of extracellular metabolites are found.

[0085] In one embodiment, the processor 10 calculates, as the optimized flux pattern, a flux pattern that minimizes the sum of the squared error between the measured and estimated values ​​of the cell proliferation rate and the squared error between the measured and estimated values ​​of the excretion amount of a specific metabolite. The measured value of the cell proliferation rate is, for example, a proliferation flux calculated based on measured proliferation data. The estimated value of the cell proliferation rate is the proliferation flux in a flux pattern that is a candidate for the optimized flux pattern. The measured value of the excretion amount of a specific metabolite is, for example, an extracellular flux of a specific metabolite calculated based on measured extracellular metabolite data. The estimated value of the excretion amount of a specific metabolite is, for example, an extracellular flux of a specific metabolite in a flux pattern that is a candidate for the optimized flux pattern.

[0086] In S212, the processor 10 performs FVA for each intracellular flux for the optimized flux pattern while maintaining the explanatory variables, cell growth rate, and excretion amount of a specific metabolite, and obtains the minimum and maximum values.

[0087] FVA is an analytical method for calculating the theoretical maximum yield and theoretical minimum yield under constraints to determine the range of flux variation. In this embodiment, the objective function is the extracellular flux of a specific metabolite. More specifically, the minimum and maximum values ​​of each intracellular flux are determined under the condition that they match the extracellular flux value of the specific metabolite calculated from the measured value. Note that this alone does not take cell proliferation capacity into consideration, so when calculating the theoretical maximum yield while maintaining a certain level of cell proliferation capacity, a constraint is also added that the cell proliferation rate must be within a predetermined percentage of the theoretical maximum.

[0088] The advantages of using FVA in this embodiment will now be described in more detail. By solving the optimization problem, one optimized flux pattern is estimated. However, there are multiple flux patterns with the same explanatory variables, cell growth rate, and target substance excretion flux, and the estimated optimized flux pattern is merely one example. Furthermore, in actual organisms, it is natural for each flux value to have a certain degree of freedom (fluctuation). Therefore, in FVA, the minimum and maximum values ​​of intracellular flux are determined, and flux patterns within that range are also considered.

[0089] Unlike FBA, the cell proliferation rate is not uniquely fixed to a maximum value, but is assumed to fall within a certain percentage of the maximum value. This is because in actual organisms, metabolic reactions that satisfy the theoretical maximum cell proliferation rate are not necessarily occurring. For example, if flux pattern A is the one that maximizes the theoretical cell proliferation rate and flux pattern B is the one that maximizes the theoretical cell proliferation rate slightly below the maximum, it is quite possible that flux pattern B is occurring within the cell.

[0090] This allows FVA to estimate multiple plausible flux patterns (primary flux patterns) with each flux having a degree of freedom within a certain range, which is thought to improve the possibility of estimating flux patterns that are in line with metabolic reactions occurring in actual cells compared to simply solving FBA.

[0091] In S213, for a flux whose degree of freedom is greater than or equal to a predetermined threshold, specifically, whose difference between the minimum and maximum values ​​is greater than or equal to a predetermined threshold, the processor 10 further estimates a primary flux pattern using FBA for each of the cases where the flux is set to the minimum value and the case where the flux is set to the maximum value.

[0092] For example, if a predetermined intracellular flux is fixed to a minimum value, the values ​​of the other fluxes are calculated based on the measured values. The minimum value of the predetermined intracellular flux and the corresponding values ​​of the other fluxes constitute a primary flux pattern.

[0093] For fluxes with little flexibility, specifically, fluxes whose values ​​barely change even when the cell growth rate is changed within a predetermined percentage, the values ​​can be fixed. In one embodiment, processor 10 uniquely fixes the values ​​of fluxes whose minimum and maximum values ​​differ by less than a predetermined threshold. This prevents unnecessary variations in the primary flux patterns, which are nearly identical.

[0094] After S213 is completed, the processor 10 advances the process to S31 in FIG. 5. The effects of the flux pattern estimation method according to this embodiment will be described below in accordance with the processes shown in FIGS. 6 and 7. According to S11 and S12, the proliferation flux and the extracellular flux can be calculated from the measured values. This makes it possible to estimate a flux pattern that matches the measured values. In particular, by using extracellular metabolome data that comprehensively measures extracellular metabolites, it is possible to estimate a flux pattern that is more realistic.

[0095] In S211, an optimized flux pattern that minimizes the error between the cell proliferation rate of the second cell line and the extracellular flux of the target substance can be simply estimated based on the measured values.

[0096] On the other hand, since there are multiple flux patterns with the same explanatory variables, cell proliferation rate, and target substance efflux, the optimized flux pattern determined in S211 can be considered just one example. By performing FVA in S212, the allowable range of each intracellular flux can be determined. This allows various flux patterns (first-order flux patterns) with the same explanatory variables, cell proliferation rate, and target substance efflux but different intracellular fluxes to be estimated in S213.

[0097] On the other hand, even if various flux patterns are obtained, presenting similar flux patterns to a plurality of users may not only not be helpful for the user's consideration but may also be cumbersome. Therefore, in S31, a representative flux pattern, which is a representative flux pattern, is extracted and presented to the user, thereby improving the efficiency of considering the factors that have caused changes in the discharge amount of the target substance.

[0098] Furthermore, in S41, the degree of agreement between the measured and estimated values ​​for the cell proliferation rate and extracellular flux among the representative flux patterns is calculated. Then, by ranking the representative flux patterns based on the degree of agreement, the user can easily determine which representative flux pattern is most plausible (highly likely to be true).

[0099] As described above, in the analysis of cellular metabolic flux distribution, it is possible to estimate multiple metabolic flux distribution patterns that are likely to match the metabolic reactions actually occurring in the cell based on the measured values ​​of substances excreted outside the cell. Furthermore, it is possible to estimate a representative pattern from among the multiple metabolic flux distribution patterns.

[0100] Furthermore, in S51, the user can identify the cause of changes in the production amount of the target product by comparing the estimated results of the representative flux patterns of the first cell line and the second cell line. If the flux pattern of the first cell line has already been estimated, the user can estimate only the representative flux pattern of the second cell line in S11 to S41 and compare it with the existing flux pattern of the first cell line in S51.

[0101] In S61, the estimated flux pattern is presented to the user, allowing the user to recognize the factors that cause changes in production volume. The user can also take specific approaches to address these factors, thereby enabling the user to control changes in production volume.

[0102] More specifically, if the change in production volume is desirable, the user can maintain the change in production volume by taking measures to keep the factor stable. Furthermore, when culturing the same cell line next time or later, the desired production volume can be reproduced by reproducing the factor. This allows the establishment of a desirable culture system.

[0103] Furthermore, based on knowledge of the factors, it is possible to explore ways to further improve production, for example, by adding substances to the medium that affect the flux, or by manipulating genes that affect the flux.

[0104] Furthermore, by clarifying the factors, it is easy for a third party with different techniques, equipment, and culture scale to reproduce the factors, and therefore, desirable changes in production volume can be easily implemented over a wider range.

[0105] On the other hand, if the change in production volume is undesirable, the user can consider measures to prevent the cause, thereby preventing a situation in which production volume deteriorates due to the same cause.

[0106] As described above, according to the method for estimating a flux pattern according to this embodiment, the stability and expandability of the culture system can be improved by controlling the culture system.

[0107] While the above example illustrates a case in which the target substance is a metabolite excreted from cells, this embodiment can also be applied to a case in which the target substance is a metabolite accumulated intracellularly and collected by disrupting cells. Metabolites accumulated intracellularly include, for example, metabolites constituting intracellular organelles or cytoskeleton. In one example, the amount of a target substance can be measured by fractionating a portion of a cell culture medium at a predetermined time (e.g., at predetermined intervals), disrupting cells from the fraction, and measuring the target substance. The extracellular flux can be measured using a portion of the culture medium in which the target substance was measured, or using another portion of the culture medium. The cell proliferation rate can be measured using a portion of the culture medium in which the target substance and / or extracellular flux was measured, or it can be measured in the original culture medium, for example, by measuring turbidity. This makes it possible to measure the amount of the target substance, the cell proliferation rate, and the extracellular flux. Therefore, the flux pattern estimation method according to this embodiment can identify metabolic pathways that cause changes in the production amount of metabolites accumulated intracellularly.

[0108] In addition, for metabolites in which a portion of the target substance is excreted outside the cell and the remainder accumulates inside the cell, the metabolic pathway that causes changes in the amount of target substance excreted outside the cell and / or the amount accumulated inside the cell can be determined by measuring the amount excreted outside the cell and the amount accumulated inside the cell separately.

[0109] Therefore, the method for estimating a flux pattern according to this embodiment can also be applied to determining the flux pattern and the flux that causes the change in a system in which at least a portion of the target substance accumulates intracellularly.

[0110] Furthermore, the target of the flux pattern estimation method according to this embodiment does not necessarily have to be a second cell line that has a different production amount of a specific metabolite from the first cell line, but may be any cell line. Therefore, the flux pattern estimation method according to this embodiment can also be applied to estimating the flux pattern of a specific cell line. In this case, in S11 of FIG. 5, excretion data is acquired for the specific cell line. Furthermore, in S12 to S41 and S61, processing is performed only for the specific cell line. S51 is not performed.

[0111] Aspects It will be appreciated by those skilled in the art that the exemplary embodiments described above are examples of the following aspects.

[0112] (Item 1) A method for estimating a metabolic flux distribution of a cell according to one embodiment includes the steps of acquiring excretion data including time series data of substances excreted outside the cell for each of a first cell line and a second cell line for comparison, calculating the excretion flux of the substance outside the cell based on the excretion data, estimating a plurality of first-order flux patterns based on the excretion flux, and clustering and outputting the plurality of first-order flux patterns based on the similarity between the plurality of first-order flux patterns.

[0113] According to the metabolic flux distribution estimation method described in paragraph 1, in analyzing the metabolic flux distribution of a cell, it is possible to estimate multiple metabolic flux distribution patterns that are likely to match the metabolic reactions actually occurring in the cell based on measured values ​​of substances excreted outside the cell. Furthermore, it is possible to cluster and output the multiple metabolic flux distribution patterns. This makes it possible to provide a method for accurately estimating metabolic fluxes in analyzing the metabolic flux distribution of a cell.

[0114] (Item 2) In the method for estimating a metabolic flux distribution according to item 1, the outputting step includes a step of estimating a representative flux pattern by clustering a plurality of primary flux patterns.

[0115] According to the method for estimating metabolic flux distribution described in Section 2, it is possible to estimate a representative pattern from among a plurality of metabolic flux distribution patterns.

[0116] (Item 3) The method for estimating metabolic flux distribution according to item 2 further comprises a step of ranking the representative flux patterns.

[0117] According to the method for estimating metabolic flux distribution described in Section 3, the user can easily determine which representative flux pattern is plausible (highly likely to be true).

[0118] (4) The method for estimating metabolic flux distribution described in 3 further includes a step of detecting differences by comparing a representative flux pattern of the first cell line with a representative flux pattern of the second cell line.

[0119] According to the metabolic flux distribution estimation method described in Section 4, the user can easily recognize the difference between the representative flux pattern of the first cell line and the representative flux pattern of the second cell line. In particular, if the first cell line and the second cell line are cells that produce different amounts of a specific metabolite, the user can recognize the factors (flux) that cause changes in the production amount. Furthermore, the user can take specific approaches to the factors. Thus, the user can control changes in the production amount. The specific metabolite is, for example, the target of the culture.

[0120] (Item 5) The method for estimating metabolic flux distribution described in item 4 further comprises a step of displaying at least one of a plurality of primary flux patterns, representative flux patterns, ranked representative flux patterns, and differences.

[0121] According to the method for estimating metabolic flux distribution described in Section 5, the user can easily understand the estimation results. Furthermore, the user can plan a verification experiment for the estimation results based on the estimation results.

[0122] (Item 6) In the method for estimating a metabolic flux distribution according to any one of items 1 to 5, the first cell line and the second cell line are cells belonging to two types of strains of the same lineage.

[0123] According to the method for estimating metabolic flux distribution described in item 6, it is possible to provide a method for correctly estimating metabolic flux in cells belonging to two types of bacterial strains of the same lineage.

[0124] (Item 7) In the method for estimating a metabolic flux distribution according to any one of Items 1 to 6, the excretion data includes cell data, which is time-series data on the amount of cells, and extracellular metabolite data, which is time-series data on the amount of extracellular metabolites, which are metabolites excreted outside the cells. The excretion flux includes proliferation flux, which is the flux of cell proliferation rate, and extracellular flux, which is the flux of excretion of extracellular metabolites.

[0125] According to the method for estimating metabolic flux distribution described in item 7, it is possible to estimate a flux pattern that is consistent with the measured values ​​of cell proliferation rate and extracellular metabolites.

[0126] (Item 8) In the method for estimating metabolic flux distribution described in Item 7, the extracellular metabolites include at least one of amino acids, vitamins, metabolites involved in amino acid metabolism, metabolites involved in nucleic acid metabolism, and metabolites involved in the TCA cycle (Tricarboxylic Acid Cycle).

[0127] According to the method for estimating metabolic flux distribution described in Section 8, a flux pattern consistent with actual measurement values ​​can be estimated based on at least one measurement value, preferably all measurement values, of amino acids, vitamins, metabolites involved in amino acid metabolism, metabolites involved in nucleic acid metabolism, and metabolites involved in the TCA cycle.

[0128] (Item 9) In the method for estimating metabolic flux distribution according to item 7 or 8, the extracellular metabolite data is calculated based on extracellular metabolome data, which is data on the totality of extracellular metabolites.

[0129] According to the method for estimating metabolic flux distribution described in item 9, it is possible to estimate a flux pattern that is highly consistent with the metabolic reactions that actually occur in cells.

[0130] (Item 10) In the method for estimating a metabolic flux distribution according to any one of items 7 to 9, the first cell line and the second cell line are cells that produce different amounts of a specific metabolite. The step of estimating a plurality of first-order flux patterns includes the steps of: estimating an optimized flux pattern by solving an optimization problem in which intracellular fluxes are explanatory variables and the error between the measured and estimated values ​​of the cell proliferation rate and the error between the measured and estimated values ​​of the excretion amount of the specific metabolite are objective functions; performing flux fluctuation analysis for each intracellular flux in the optimized flux pattern while maintaining the explanatory variables, the cell proliferation rate, and the excretion amount of the specific metabolite; and, for a flux whose difference between the minimum and maximum values ​​is equal to or greater than a predetermined threshold, further estimating a first-order flux pattern using flux balance analysis for each of the cases in which the flux is set to the minimum value and the case in which the flux is set to the maximum value.

[0131] According to the method for estimating metabolic flux distribution described in Section 10, by combining the optimization problem and FVA, it is possible to estimate multiple first-order flux patterns that match the actual measured values, thereby making it possible to estimate the difference in flux that causes the difference in the production amount of a specific metabolite.

[0132] (Item 11) In the method for estimating metabolic flux distribution described in any one of items 1 to 10, the step of clustering and outputting further includes a step of clustering multiple primary flux patterns based on the similarity between the excretion fluxes.

[0133] In the method for estimating metabolic flux distribution described in paragraph 11, differences in intracellular fluxes have no effect on excretion fluxes and are therefore not important factors in the changes in excretion products. Therefore, the possibility that factors that are not important factors in the changes in excretion amounts of excretion products affect clustering can be reduced. In particular, when the excretion products include the target product of the culture, the possibility that factors that are not important factors in the changes in productivity of the target product affect clustering can be reduced.

[0134] (Item 12) In the method for estimating metabolic flux distribution described in item 3, the ranking step includes a step of ranking representative flux patterns based on the error between the measured value of excretion flux and the estimated value by flux balance analysis.

[0135] According to the method for estimating metabolic flux distribution described in Section 12, it is possible to rank the excretion flux based on the consistency between the measured value and the FBA estimated value.

[0136] (Item 13) A method for estimating a metabolic flux distribution of a cell according to a second aspect includes the steps of: acquiring excretion data for a predetermined cell line, the excretion data including time-series change data of a substance excreted outside the cell; calculating the excretion flux of the substance outside the cell based on the excretion data; estimating a plurality of first-order flux patterns based on the excretion flux; and clustering and outputting the plurality of first-order flux patterns based on the similarity between the plurality of first-order flux patterns.

[0137] According to the metabolic flux distribution estimation method described in paragraph 13, in analyzing the metabolic flux distribution of a cell, it is possible to estimate multiple metabolic flux distribution patterns that are likely to match the metabolic reactions actually occurring in the cell based on measured values ​​of substances excreted outside the cell. Furthermore, the multiple metabolic flux distribution patterns can be clustered and output. This provides a method for accurately estimating metabolic fluxes in analyzing the metabolic flux distribution of a cell.

[0138] (Item 14) An analysis device according to a third aspect is an analysis device that estimates a metabolic flux distribution of a cell. The analysis device includes a memory and a processor. The memory stores excretion data including time-series change data of substances excreted outside the cell for each of a first cell line and a second cell line as a comparison target. The processor calculates the excretion flux of the substance to the cell based on the excretion data, estimates multiple first-order flux patterns based on the excretion flux, and clusters and outputs the multiple first-order flux patterns based on the similarity between the multiple first-order flux patterns.

[0139] According to the analysis device described in paragraph 14, in analyzing the metabolic flux distribution of a cell, it is possible to estimate a plurality of metabolic flux distribution patterns that are likely to match the metabolic reactions actually occurring in the cell based on the measured values ​​of substances excreted outside the cell. Furthermore, it is possible to cluster the plurality of metabolic flux distributions and output them. This makes it possible to provide a method for accurately estimating metabolic fluxes in analyzing the metabolic flux distribution of a cell.

[0140] (Item 15) A program that, when executed by a computer, causes the computer to implement the metabolic flux distribution estimation method described in any one of items 1 to 13.

[0141] The embodiments disclosed herein should be considered to be illustrative in all respects and not restrictive. The scope of the present invention is defined by the claims, not by the above description, and is intended to include all modifications within the meaning and scope of the claims.

[0142] REFERENCE SIGNS LIST 1 Analysis device, 2 Measurement device, 3 Culture device, 10 Processor, 11 Memory, 12 Communication unit, 13 Input / output unit, 14 Operation unit, 15 Display, 19 Controller, 21 Cell measurement unit, 22 Extracellular metabolite measurement unit, 100 Analysis system.

Claims

1. A method for estimating a metabolic flux distribution in a cell, comprising: acquiring excretion data including time series data of substances excreted outside the cells for each of the first cell line and the second cell line for comparison; calculating an excretion flux of the substance out of the cell based on the excretion data; estimating a plurality of primary flux patterns based on the emission fluxes; and clustering and outputting the plurality of primary flux patterns using information about the mutual similarity of the plurality of primary flux patterns; A method for estimating metabolic flux distribution, wherein the excretion flux includes a proliferation flux, which is a flux of a cell proliferation rate, and an extracellular flux, which is a flux of excretion of extracellular metabolites, which are metabolites excreted outside the cell.

2. The method for estimating a metabolic flux distribution according to claim 1 , wherein the step of outputting includes a step of estimating a representative flux pattern by clustering the plurality of first-order flux patterns.

3. The method for estimating a metabolic flux distribution according to claim 2 , further comprising the step of ranking the representative flux patterns.

4. The method for estimating a metabolic flux distribution according to claim 3 , further comprising a step of detecting differences by comparing the representative flux pattern of the first cell line with the representative flux pattern of the second cell line.

5. The method for estimating a metabolic flux distribution according to claim 4 , further comprising the step of outputting at least one of the plurality of primary flux patterns, the representative flux pattern, the ranked representative flux pattern, and the dissimilarity point.

6. The method for estimating a metabolic flux distribution according to claim 1 , wherein the first cell line and the second cell line belong to two types of strains of the same lineage, respectively.

7. The method for estimating a metabolic flux distribution according to claim 1 , wherein the excretion data includes cell data, which is time series data on the amount of cells, and extracellular metabolite data, which is time series data on changes in the amount of the extracellular metabolites.

8. 8. The method for estimating a metabolic flux distribution according to claim 7, wherein the extracellular metabolites include at least one of amino acids, vitamins, metabolites involved in amino acid metabolism, metabolites involved in nucleic acid metabolism, and metabolites involved in the tricarboxylic acid cycle (TCA cycle).

9. The method for estimating a metabolic flux distribution according to claim 7 , wherein the extracellular metabolite data is calculated based on extracellular metabolome data, which is data on the totality of the extracellular metabolites.

10. The first cell line and the second cell line are cells that produce different amounts of a specific metabolite, The step of estimating a plurality of primary flux patterns comprises: A step of estimating an optimized flux pattern by solving an optimization problem in which the intracellular flux is used as an explanatory variable, and the error between the measured value and the estimated value of the cell proliferation rate and the error between the measured value and the estimated value of the excretion amount of the specific metabolite are used as objective functions; performing a flux fluctuation analysis for each intracellular flux in the optimized flux pattern while maintaining the explanatory variables, the cell proliferation rate, and the excretion amount of the specific metabolite, to obtain a minimum value and a maximum value; The method for estimating a metabolic flux distribution according to claim 9, further comprising the step of: for a flux in which a difference between the minimum value and the maximum value is equal to or greater than a predetermined threshold, estimating the first-order flux pattern using flux balance analysis for each of the cases in which the flux is set to the minimum value and the maximum value.

11. The method for estimating a metabolic flux distribution according to claim 1 , wherein the step of clustering and outputting further comprises the step of clustering the plurality of primary flux patterns based on a similarity between the excretion fluxes.

12. The method for estimating a metabolic flux distribution according to claim 3 , wherein the ranking step includes a step of ranking the representative flux patterns based on an error between a measured value of the excretion flux and an estimated value by flux balance analysis.

13. A method for estimating a metabolic flux distribution in a cell, comprising: acquiring excretion data including time series change data of a substance excreted outside the cell for a predetermined cell line; calculating an excretion flux of the substance out of the cell based on the excretion data; estimating a plurality of primary flux patterns based on the emission fluxes; and clustering and outputting the plurality of primary flux patterns based on a similarity between the plurality of primary flux patterns, A method for estimating metabolic flux distribution, wherein the excretion flux includes a proliferation flux, which is a flux of a cell proliferation rate, and an extracellular flux, which is a flux of excretion of extracellular metabolites, which are metabolites excreted outside the cell.

14. An analysis device for estimating a metabolic flux distribution of a cell, comprising: a memory for storing excretion data including time series change data of a substance excreted outside the cells for each of the first cell line and the second cell line for comparison; a processor; The processor, Calculating an excretion flux of the substance out of the cell based on the excretion data; estimating a plurality of primary flux patterns based on the emission fluxes; clustering the plurality of primary flux patterns based on the mutual similarities of the plurality of primary flux patterns and outputting the clustered results; The efflux flux includes a proliferation flux, which is a flux of a cell proliferation rate, and an extracellular flux, which is a flux of efflux of extracellular metabolites, which are metabolites that are effluxed outside the cell.

15. A program that, when executed by a computer, causes the computer to carry out the metabolic flux distribution estimation method according to claim 1.