Culture system, culture method, and culture management device

Abstract

Provided is a culture system making it possible to reduce the time and cost needed for culture process development. A culture system (1) is provided with: a computation unit (12) for outputting additional experimental information that is information on an explanatory variable to modify in a culture experiment on the basis of values of a plurality of explanatory variables; a plurality of culture tanks (30); and a control unit (13) for controlling transfer of cells to be cultured between the plurality of culture tanks (30), and the like, on the basis of the additional experimental information outputted by the computation unit (12). The plurality of explanatory variables include: one or more first explanatory variables the value of which can be modified in a predetermined range for each of the plurality of culture vessels (30); and at least one second explanatory variable the possible value of which is varied among at least a set of culture tanks (30) included in the plurality of culture tanks (30). If the explanatory variable to be modified is the second explanatory variable, the control unit (13) determines whether the cells to be cultured need to be transferred to a second culture tank (30b) different from a first culture tank (30a), and controls the transfer of the cells to be cultured.

Description

Cultivation system, culture method, and culture management device

[0001] The present invention relates to a culture system, a culture method, and a culture management device.

[0002] Methods for producing useful substances by culturing cells of plants, microorganisms, animals, etc. are used in various industries, including brewing, food, chemicals, and pharmaceuticals. For example, biopharmaceuticals, such as antibody drugs, contain substances produced by animal cells as their main components, and these substances can be obtained by culturing animal cells and isolating and purifying the target substances secreted into the culture medium.

[0003] When constructing a production process for the above-mentioned useful substances, first, process conditions such as culture conditions such as temperature and pH are optimized in a small scale (a few mL to several tens of liters) in laboratory experiments. Then, in constructing the production process, the process is scaled up to the desired large scale (hundreds of liters, thousands of liters or more) so as to maintain the production efficiency and product quality of the process optimized in the laboratory experiments. Known scale-up methods include scale-up based on the law of similarity and scale-up using design space.

[0004] Scaling up based on the law of similarity is generally performed based on geometric similarity (all dimensional ratios of the structure and layout of the mixing equipment are similar) and on the criteria of maintaining a constant mixing power per unit liquid volume. Specifically, the vessel is designed so that the mixing power required, power consumption per unit liquid volume, mixing impeller rotation speed, mixing impeller diameter, liquid discharge rate, liquid circulation rate within the reactor, mixing impeller tip speed, and Reynolds number are constant before and after scaling up. However, since it is difficult to scale up in a way that satisfies all of these criteria simultaneously, scaling up is performed by focusing on satisfying the most important mixing criteria (items that affect productivity and quality).

[0005] Here, Figure 14 is an explanatory diagram illustrating the concept of scale-up of a fermentor using design space, which is one of the conventional techniques. In scale-up using design space, as shown in the left diagram of Figure 14, the culturable range is evaluated in a laboratory experiment, centered on the optimal operating conditions (optimum point) obtained in a small-scale laboratory experiment. Then, as shown in the right diagram of Figure 14, when designing a desired large-scale production-scale tank, the tank is designed so that a culturable range exists through fluid analysis. Generally, the culturable range in a large-scale production-scale tank is narrower than the culturable range in a small-scale tank.

[0006] Technologies that can be used to build the above-described production process are described, for example, in Patent Documents 1 and 2. Patent Document 1 describes a culture-related process optimization method for optimizing a culture-related process related to cell culturing. This method includes an acquisition step, a variable parameter item identification step, an execution procedure generation step, an execution result acquisition step, an evaluation result acquisition step, and a storage step. The acquisition step acquires a starting execution procedure that defines the operations performed in the culture-related process and serves as a starting point for the search. The variable parameter item identification step identifies variable parameter items in the starting execution procedure for which variable parameter values ​​can be set. The execution procedure generation step generates an execution procedure by setting variable parameter values ​​for the variable parameter items based on past execution results and their evaluation results. The execution result acquisition step acquires the execution results when an execution entity actually executes the execution procedure in an execution environment. The evaluation result acquisition step acquires evaluation results for the execution results using an evaluation result acquisition unit. The storage step associates and records the execution procedure, variable parameter values, execution results, and evaluation results.

[0007] Patent Document 2 also describes a cell culture chamber for automating the production of antigen-specific T cells. The chamber has a bottom surface made of a first material to which cells adhere. The chamber has at least one additional surface made of a gas-permeable second material. The chamber has one or more inlet ports. The chamber also has one or more outlet ports. The one or more inlet ports and the one or more outlet ports are arranged to move fluid within the cell culture chamber at least partially along a flow path perpendicular to the bottom surface. Patent Document 2 also describes that the chamber has a surface area comparable to that of conventional well plates, such as 6-well plates and 24-well plates. Patent Document 2 also describes that the height of the cell culture chamber is, for example, from 0.5 mm to 100 mm.

[0008] JP 2023-35238 A JP 2019-522980 A

[0009] In the process development for the above-mentioned production, as mentioned above, the culture conditions are first optimized on a small scale. However, there are many items to consider, such as temperature, pH, dissolved oxygen (DO) concentration, and oxygen aeration rate, and optimization requires numerous experiments, which are very costly and time-consuming. Considering the commercialization of useful substances, sufficient optimization may not be achieved, and process development may end with the minimum culture conditions necessary for commercialization.

[0010] Furthermore, when tank design is performed using the scale-up method, numerous small-scale culture experiments are conducted to optimize the process. However, when tank design is performed using the scale-up method, the culture experiment data required for scale-up must be collected again for each scale-up, which tends to lengthen the time required to build a production process from a small scale.

[0011] Furthermore, small-scale culture experiments broaden the search range for specific culture condition items (explanatory variables). Therefore, it is not possible to build a design space using a single type of culture device; two or more types of culture devices with different configurable ranges are required. For example, to evaluate shear stress due to stirring, multiple culture devices with different impeller shapes and rotation speeds, such as low shear stress range and high shear stress range, are required. In this case, when conducting optimization experiments using machine learning, operations such as replacing the impellers during the culture are required, making it difficult to optimize culture conditions that require culture in a sterile environment.

[0012] The culture-related process optimization method described in Patent Document 1 allows for optimization of culture conditions using machine learning, but does not allow for switching of culture equipment based on explanatory variables. Therefore, this method may have difficulty optimizing culture conditions that require culture in a sterile environment. Furthermore, the cell culture chamber described in Patent Document 2 automates the production of antigen-specific T cells and does not take into account scaling up from small to large scales. Furthermore, while the cell culture chamber described in Patent Document 2 is configured to allow for the transfer of culture medium, it is considered difficult to apply machine learning related to scale-up, such as multi-objective optimization. Therefore, even considering Patent Documents 1 and 2, conventional technologies have not been able to sufficiently reduce the time and cost required for culture process development.

[0013] The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a culture system, a culture method, and a culture management device that can reduce the time and cost required for developing a culture process.

[0014] In order to solve the above problems, the present invention provides culture optimization in a culture system that includes two or more culture tanks with different settable ranges for a specific culture condition item (explanatory variable), selects the culture tank that is suitable for the culture conditions proposed by an optimization tool, and has equipment and a control mechanism for transferring a culture solution containing cells to the selected culture tank.The details are as follows.

[0015] The culture system according to the present invention, which has solved the above-mentioned problems, is a culture system for conducting experiments to search for culture conditions in cell culture, and includes a calculation unit that searches for values ​​of a plurality of explanatory variables that will cause a target variable in the culture experiment to satisfy a predetermined condition based on experimental data from the culture experiment, in which a plurality of explanatory variables related to the culture conditions are changed, and outputs additional experimental information, which is information on the explanatory variables to be changed in the culture experiment, based on the searched values ​​of the plurality of explanatory variables; a plurality of culture vessels for the culture experiment; and controls the culture conditions in the plurality of culture vessels based on the additional experimental information output by the calculation unit, and controls the transfer of cells to be cultured between the plurality of culture vessels in accordance with the determination. and a control unit configured to control the transfer of the cells to be cultured from the first culture tank to the second culture tank when the explanatory variable to be changed in the culture experiment performed in a first culture tank of the plurality of culture tanks in the additional experimental information is the second explanatory variable, and when the result indicates that the cells to be cultured need to be transferred, the control unit controls the transfer of the cells to be cultured from the first culture tank to the second culture tank.

[0016] According to the present invention, it is possible to provide a culture system, a culture method, and a culture management device that can reduce the time and cost required for culture process development. Problems, configurations, and effects other than those described above will become clear from the description of the following embodiments.

[0017] 9A is a diagram illustrating an overview of the culture system 1 according to the present embodiment. FIG. 9B is a flow chart illustrating an optimization process by machine learning. FIG. 9C is a schematic diagram illustrating an example of the configuration of a general culture device 300. FIG. 9D is an explanatory diagram illustrating a response when the settable range of explanatory variables is narrow. FIG. 9E is a configuration diagram illustrating an example of the configuration of multiple culture tanks 30 used in the culture system 1. FIG. 9F is an explanatory diagram illustrating a state before cells to be cultured are transferred from a first culture tank 30a to a second culture tank 30b. FIG. 9G is an explanatory diagram illustrating a state during which cells to be cultured are being transferred from the first culture tank 30a to the second culture tank 30b. FIG. 9H is an explanatory diagram illustrating a state after cells to be cultured have been transferred from the first culture tank 30a to the second culture tank 30b. FIG. 9H is a graph of a function used in a test calculation in multi-objective optimization, the graph being expressed in three dimensions. FIG. 9G is a graph of a function used in a test calculation in multi-objective optimization, the graph being expressed in two dimensions. FIG. 9H is a graph of a function used in a test calculation in multi-objective optimization, the graph being expressed in three dimensions. 9C is a graph of a function used in a test calculation in multi-objective optimization, and is a two-dimensional representation of the graph in FIG. 9C. FIG. 9C is a graph showing an example of a Pareto front for the objective variable M (e.g., productivity) and the objective variable N (e.g., quality). FIG. 9D is a flow diagram explaining the contents of a culture method according to one embodiment of the present invention. FIG. 9C is an explanatory diagram explaining an overview of a culture management device 120. FIG. 9D is a graph showing the relationship between the number of culture days (days) and the antibody concentration under conditions before optimization (conditions for the first cycle) and conditions after optimization (culture conditions for the 20th cycle). FIG. 9D is an explanatory diagram explaining the concept of scaling up a culture tank using design space, which is one of the prior art techniques.

[0018] In order to achieve the above-mentioned objects, the present invention has the following configuration. Note that the objects, features, advantages, and ideas of the present invention will be apparent to those skilled in the art from the description in this specification, and those skilled in the art will be able to easily reproduce the present invention from the description in this specification. The specific embodiments of the invention described below show preferred embodiments of the present invention and are shown for illustrative or explanatory purposes, and are not intended to limit the present invention thereto. It will be apparent to those skilled in the art that various changes and modifications can be made based on the description in this specification within the spirit and scope of the present invention disclosed in this specification.

[0019] An embodiment of the present invention will be described below with reference to the drawings. However, the embodiment described below is an example for explaining the present invention, and appropriate omissions and simplifications have been made for clarity of explanation. The present invention can be implemented in various other forms. Unless otherwise specified, each component may be singular or plural. The position, size, shape, range, etc. of each component shown in the drawings may not represent the actual position, size, shape, range, etc., in order to facilitate understanding of the invention. Therefore, the present invention is not necessarily limited to the position, size, shape, range, etc. disclosed in the drawings.

[0020] When there are multiple components with the same or similar functions, they may be described using the same reference numeral with different subscripts. Also, when there is no need to distinguish between these multiple components, the subscripts may be omitted. Components with the same or similar functions may be described using the same reference numeral, and duplicate descriptions may be omitted.

[0021] In the embodiments, processing performed by executing a program may be described. Here, a computer executes the program using a processor (e.g., a CPU or a GPU) and performs processing defined by the program using storage resources (e.g., memory) and interface (I / F) devices (e.g., communication ports). Therefore, the entity performing the processing by executing the program may be the processor. Similarly, the entity performing the processing by executing the program may be a controller, device, system, computer, or node having a processor. The entity performing the processing by executing the program may be any computing unit, and may include a dedicated circuit that performs specific processing. Here, the dedicated circuit may be, for example, an FPGA (Field Programmable Gate Array), an ASIC (Application Specific Integrated Circuit), or a CPLD (Complex Programmable Logic Device).

[0022] A program may be installed on a computer from a program source. The program source may be, for example, a program distribution server or a computer-readable storage medium. When the program source is a program distribution server, the program distribution server may include a processor and storage resources for storing the program to be distributed, and the processor of the program distribution server may distribute the program to be distributed to other computers. In addition, in an embodiment, two or more programs may be realized as one program, or one program may be realized as two or more programs.

[0023] <Culture System> A culture system according to one embodiment of the present invention relates to a data collection method and a small-scale culture device for data collection, which are used to establish a manufacturing process for mass-culturing cells for the production of useful substances, for example. First, an overview of one embodiment of the culture system 1 according to the present invention will be described with reference to FIG. 1. FIG. 1 is an explanatory diagram illustrating an overview of the culture system 1 according to this embodiment. The culture system 1 shown in FIG. 1 performs experiments to explore culture conditions for cell culture.

[0024] 1, the culture system 1 includes a calculation unit 12, a plurality of culture vessels 30, and a control unit 13. The calculation unit 12 and the control unit 13 are components of a computer 10. The calculation unit 12 and the control unit 13 are configured, for example, by a central processing unit (CPU 11) that executes a program that controls the operation of each configuration, function, and processing unit.

[0025] That is, as shown in FIG. 1 , the culture system 1 includes a computer 10. The computer 10 includes a CPU 11, a RAM 14, a ROM 15, a HDD 16, a communication I / F 17, an input / output I / F 18, and a media I / F 19. The communication I / F 17 is connected to a communication circuit 20. The input / output I / F 18 is connected to an input / output device 21. The media I / F 19 reads and writes data from a recording medium 22. The ROM 15 stores application programs executed by the CPU 11, various data, and the like. The HDD 16 stores experimental data, various data, control programs, and the like as appropriate. The CPU 11 executes the application programs loaded into the RAM 14 to realize various functions. The culture system 1 may be configured using a single computer 10, or may be configured using multiple computers 10 connected via the communication circuit 20.

[0026] In this embodiment, the calculation unit 12 searches for values ​​of multiple explanatory variables that will cause the objective variable in the culture experiment to satisfy a predetermined condition, based on experimental data from a culture experiment conducted by changing multiple explanatory variables related to the culture conditions. The calculation unit 12 also outputs additional experimental information, which is information on the explanatory variables to be changed in the culture experiment, based on the searched values ​​of the multiple explanatory variables. The search for values ​​of the multiple explanatory variables can be performed using machine learning, specifically Bayesian optimization or multi-objective optimization. The search for values ​​of the multiple explanatory variables and machine learning will be described later. The objective variable and explanatory variables will also be described later.

[0027] Furthermore, multiple culture tanks 30 are used in the culture experiment described above. The multiple culture tanks 30 include, for example, multiple first culture tanks 30a and multiple second culture tanks 30b. The second culture tanks 30b may be culture tanks separate from the first culture tank 30a, and the number of second culture tanks 30b is not particularly limited. For example, FIG. 1 illustrates a state in which one second culture tank 30b is connected to one first culture tank 30a, but two or more second culture tanks 30b may be connected. In this case, the first culture tank 30a and two or more second culture tanks 30b may be connected in series, such as first culture tank 30a-second culture tank 30b-second culture tank 30b, or may be connected to the first culture tank 30a in the order of second culture tank 30b-first culture tank 30a-second culture tank 30b. The multiple culture tanks 30 can all be small-scale with a volume of several mL to several tens of liters.

[0028] The control unit 13 controls the culture conditions in the plurality of culture tanks 30 based on the additional experimental information output by the calculation unit 12, and also controls the transfer of cells to be cultured between the plurality of culture tanks 30 based on the judgment described below.

[0029] Here, the plurality of explanatory variables include one or more first explanatory variables whose values ​​can be changed within a predetermined range in each of the plurality of culture tanks 30. That is, the first explanatory variable is a variable whose value can be changed within the same predetermined numerical range (for example, the same predetermined numerical range of 1 to 100) in all of the plurality of culture tanks 30. Furthermore, the plurality of explanatory variables include one or more second explanatory variables whose possible values ​​differ between at least one set of culture tanks 30 included in the plurality of culture tanks 30 (specifically, a set of a first culture tank 30a and a second culture tank 30b (including cases where there are multiple second culture tanks 30b)). That is, the second explanatory variable has different possible values ​​between a set of culture tanks 30, for example, one (first culture tank 30a) has a value in the range of 1 to 40, the other (second culture tank 30b) has a value in the range of 35 to 75, and if there is another one, the other (second culture tank 30b) has a value in the range of 70 to 100.

[0030] Then, when the explanatory variable to be changed in the culture experiment to be performed in the first culture tank 30a among the plurality of culture tanks 30 in the additional experiment information is the second explanatory variable, the control unit 13 determines whether or not it is necessary to transfer the cells to be cultured to a second culture tank 30b, which is different from the first culture tank 30a, among the plurality of culture tanks 30. When the result indicates that transfer is necessary, the control unit 13 controls the transfer of the cells to be cultured from the first culture tank 30a to the second culture tank 30b. Note that examples of cases where the result indicates that transfer is necessary include when the conditions (numerical range) of the second culture tank 30b do not fall within the conditions (numerical range) of the first culture tank 30a, or when control becomes difficult.

[0031] Here, a method for searching for suitable values ​​for multiple explanatory variables in the culture system 1 having the above configuration, that is, an optimization process using machine learning, will be described. FIG. 2 is a flow diagram illustrating the optimization process using machine learning. As shown in FIG. 2, the optimization process using machine learning consists of [1] experimental condition setting, [2] culture experiment, [3] analysis, [4] input to the calculation unit (including recording), [5] calculation for proposing the next experimental condition, [6] presentation of the next experimental condition, and [7] termination condition determination. The process is performed in order from [1], and if the termination condition is not met in [7], the cycle returns to [1] and repeats. Each component will be described below.

[0032] [1] Experimental Condition Setting: In the optimization of culture conditions according to this embodiment, optimal conditions are reached by repeating the cycles [1] to [7] above. In the first cycle of experiments, experimental conditions can be set arbitrarily. The experimental conditions can be standard culture conditions found in literature or one of the culture conditions previously used. Furthermore, if results under multiple culture conditions are already available, the data can be entered in "[4] Input to Calculation Unit (including recording)" (described below) and used as learning data for the calculation in [5]. Then, depending on the learning results, culture conditions can be entered in "[6] Presentation of Next Experimental Conditions." The culture experiment results to be learned in this case can be a set of data obtained in previous developments, or the results of individual culture experiments for a group of experimental conditions evenly distributed within a predetermined search range can be obtained and used as learning data. The latter learning data is desirable because it enables efficient search for the optimal point in "[5] Calculation for Proposing Next Experimental Conditions," thereby reducing the number of cycles.

[0033] [2] Culture Experiments Culture experiments require a mechanism that allows the optimization of culture conditions (explanatory variables) and analyzes the objective variables. Specific examples of explanatory variables and objective variables are listed below, taking the scale-up of the culture tank 30 as an example. However, they are not limited to the items listed here.

[0034] 1. Selection of Objective Variables and Explanatory Variables The objective variable is the productivity and quality of the substance to be produced. Examples of objective variables include, but are not limited to, the concentration of the product, the total yield of the product, and the purity of the product. The objective variable does not change before and after scale-up. The explanatory variables are divided into variables that do not change before and after scale-up (the first explanatory variables described above) and variables that can be set and controlled before scale-up but are not set or controlled after scale-up (the second explanatory variables described above). The first explanatory variables are parameters that are set and controlled after scale-up, and examples include temperature, pH, and DO concentration. Examples of the second explanatory variables include shear stress, bubble diameter, Kolmogorov scale, and nutrient concentration distribution. At least one of these can be selected as the second explanatory variable. Setting these as the second explanatory variables allows for appropriate search and optimization within a wide range of variables that are not set or controlled after scale-up.

[0035] FIG. 3 is a schematic diagram illustrating an example of the configuration of a typical culture apparatus 300. This culture apparatus 300 is a culture apparatus for perfusion culture. The culture apparatus 300 is composed of a culture tank 301, a mechanism 302 for controlling the addition of medium components by feeding, a mechanism for removing the cell-containing culture solution by bleeding (not shown), a cell separation mechanism 303 for recovering target substances while leaving the cells in the culture tank 301, and a mechanism (not shown) for controlling the temperature, pH, dissolved oxygen, and mixability within the culture tank 301. In perfusion culture, appropriate control of feeding, cell separation, and bleeding can maintain steady states for not only temperature, pH, dissolved oxygen, and mixability but also nutrient concentrations, waste product concentrations, and cell density. Perfusion culture is advantageous because it provides higher quality data than conventional batch and fed-batch cultures. In batch culture and fed-batch culture, fluctuations in cell density occur due to cell consumption of nutrients and accumulation of waste products, as well as cell growth. However, if these fluctuations are within acceptable limits, culture equipment for batch culture and fed-batch culture can also be used.

[0036] Among the explanatory variables in the general culture apparatus 300 (perfusion culture, batch culture, fed-batch culture), temperature and pH can be set and controlled over a relatively wide range. However, when the explanatory variables are shear stress or aeration bubble diameter, the range of values ​​that can be set is limited by the shape of the agitator blades and the pore size of the sparger used for aeration. Therefore, for example, when shear stress is used as the explanatory variable, pitched paddles are used in low shear stress regions, flat paddles in medium shear stress regions, and disc turbine blades in high shear stress regions, and blade replacement is required depending on the culture conditions. Regarding aeration, spargers with different pore sizes are also replaced. However, the culture solution during cultivation must be protected from contamination by external bacteria, making it a closed system. Therefore, replacing the agitator blades or sparger during cultivation is not appropriate.

[0037] In consideration of this, the culture system 1 according to this embodiment uses multiple culture tanks 30 with different settable ranges for a specific explanatory variable (see FIGS. 1 and 4). FIG. 4 is an explanatory diagram illustrating how to handle a case where the settable range of the explanatory variable is narrow. As shown in FIG. 4, the culture system 1 has culture tanks 30 corresponding to conditions A (a≦x≦b), B (c≦x≦d), and C (e≦x≦f), respectively, when controlling the second explanatory variable. The culture system 1 switches between culture tanks 30 appropriate for the second explanatory variable indicated by x. Among the aforementioned a to f, the magnitude relationships between a to e are a≦c≦b and c≦e≦d. In this example, the numerical ranges for conditions A to C are set to overlap. While FIG. 4 illustrates the second explanatory variable, the first explanatory variable can be handled in a similar manner.

[0038] In the case of this embodiment, for example, as conditions A to C described above, a culture tank 30 equipped with at least one of pitched paddles, flat paddles, and disk turbine blades as stirring blades for controlling shear stress may be prepared, and a culture tank 30 may be appropriately selected from these based on the judgment of the control unit 13. In this way, the shear stress, which is a variable that is set and not controlled after scale-up, can be appropriately searched and optimized within a wide range. This can also be achieved by preparing multiple culture tanks 30 equipped with spargers with different pore sizes, which allows appropriate search and optimization of the bubble diameter within a similarly wide range.

[0039] For example, when a suitable explanatory variable value is presented for the second fermenter 30b in the Nth cycle from the first fermenter 30a used in the N-1th cycle (for example, when an instruction to change from condition A to condition B is given), the culture system 1 transfers the cell-containing culture medium from the first fermenter 30a to the second fermenter 30b. In this embodiment, as shown in FIG. 4, when a wide range of optimization conditions are searched for for a specific explanatory variable, multiple fermenters 30 with different settable ranges are provided, and when "(6) Presentation of the next experimental conditions" is performed, it is determined which fermenter 30 should be used for the specific explanatory variable. The number of explanatory variables for the characteristics is not limited to one, and the same applies when there are multiple explanatory variables.

[0040] FIG. 5 is a configuration diagram illustrating an example of the configuration of the multiple culture tanks 30 used in the culture system 1. As described above, the multiple culture tanks 30 have different settable ranges for a specific explanatory variable. As shown in FIG. 5, the multiple culture tanks 30 preferably include a pipe 31a that supplies gas to the first culture tank 30a and a gas supply valve 31 provided thereon. The multiple culture tanks 30 preferably include a pipe 32c that connects the first culture tank 30a and the second culture tank 30b and transfer valves 32a and 32b provided thereon. The multiple culture tanks 30 preferably include a pipe 33a that discharges gas from the second culture tank 30b and a gas discharge valve 33 provided thereon. The gas supplied via the gas supply valve 31 is preferably sterilized using an air filter or the like. When multiple culture tanks 30 are configured in this manner, when transferring cells to be cultured from the first culture tank 30a to the second culture tank 30b, the control unit 13 first opens the gas supply valve 31 to start supplying gas to the first culture tank 30a. The control unit 13 then opens the transfer valves 32a, 32b and the gas exhaust valve 33 to transfer the cells to be cultured. In this manner, the culture medium CM containing cells can be transferred from the first culture tank 30a to the second culture tank 30b in an appropriate and prompt manner while maintaining a sterile state using the gas pressure. The multiple culture tanks 30 are not limited to the embodiment shown in FIG. 5 , as long as an appropriate culture tank 30 can be selected each time "(6) Presentation of Next Experimental Conditions" is performed.

[0041] As shown in FIG. 5 , the multiple culture tanks 30 preferably include a feed mechanism 34, a cell separation mechanism 35, a bleeding mechanism 36, and temperature, pH, and dissolved oxygen control mechanisms (not shown). The feed mechanism 34 adds medium components by feeding. The cell separation mechanism 35 recovers target substances while leaving cells in the culture tank 30. The bleeding mechanism 36 removes the culture solution CM containing cells by bleeding. The temperature, pH, and dissolved oxygen control mechanism controls the temperature, pH, and dissolved oxygen. These mechanisms are provided with valves in the connecting pipes to control the flow rates of gases, culture solution CM, etc., as needed. By having these configurations, the multiple culture tanks 30 can be used to perform culture in any of the culture tanks 30.

[0042] An example of a method and apparatus for transferring cells to be cultured from a first culture tank 30a to a second culture tank 30b is shown in Figures 6 to 8. Figure 6 is an explanatory diagram showing the state before the cells to be cultured are transferred from the first culture tank 30a to the second culture tank 30b. Figure 7 is an explanatory diagram showing the state during the transfer of the cells to be cultured from the first culture tank 30a to the second culture tank 30b. Figure 8 is an explanatory diagram showing the state after the cells to be cultured have been transferred from the first culture tank 30a to the second culture tank 30b.

[0043] As shown in FIG. 6, before transfer, all valves (gas supply valve 31, transfer valves 32a, 32b, and gas exhaust valve 33) of the first fermentation tank 30a and the second fermentation tank 30b are closed. Next, as shown in FIG. 7, the gas supply valve 31 on the first fermentation tank 30a side (before transfer) is opened, gas is vented into the first fermentation tank 30a, and the internal pressure of the first fermentation tank 30a is increased. Next, the transfer valves 32a, 32b and the gas exhaust valve 33 are opened. This starts the transfer of the culture solution CM to the second fermentation tank 30b. Then, as shown in FIG. 8, after the culture solution CM has completely transferred to the second fermentation tank 30b, all valves (gas supply valve 31, transfer valves 32a, 32b, and gas exhaust valve 33) are closed. It is desirable that the first fermentation tank 30a, which has been emptied by this transfer, be provided with a mechanism for aseptically cleaning the inside of the first fermentation tank 30a. For example, a sterile buffer solution is added from the outside, stirred in the first culture tank 30a, and then discharged outside the first culture tank 30a.

[0044] When the process is optimized for scale-up using the first fermenter 30a and the second fermenter 30b described above, data is collected through a culture experiment as follows.

[0045] 2. Data collection through culture experiments A culture experiment is performed using the cells to be used, with the first and second explanatory variables selected in "1. Selection of Objective Variable and Explanatory Variable" above. The first explanatory variable is varied as in the past, with each parameter varied. The objective variable under those conditions is then measured. The second explanatory variable is a parameter that has not been measured in small-scale fermentors in the past for optimizing the culture process. Examples of measurement methods for each second explanatory variable and the device functions for measurement are described below. Examples of second explanatory variables and data collection using them are shown in Figure 4.

[0046] Shear stress: Shear stress is distributed within the culture tank 30 and is difficult to measure directly. Therefore, the shear stress distribution when the agitation rotation speed is changed is quantified by fluid analysis (simulation) based on the structure of the culture tank 30 used in the culture experiment. Conventionally, inclined paddle-type agitation blades are often used, but in this case, even at the maximum rotation speed, the magnitude of the shear stress after scale-up is not reached, so it is desirable to increase the diameter of the agitation blades and use vertically long flat paddles or disk turbine blades.

[0047] Nutrient concentration: Nutrient concentration is controlled by measuring the nutrient concentration in the culture medium CM and adding a nutrient-containing medium to achieve a set value based on that value. Examples of nutrients include glucose, glutamine, and glutamic acid. Nutrient concentration can be measured by sampling the culture medium CM and using high-performance liquid chromatography (HPLC) or liquid chromatography mass spectrometry (LC-MS), or by in-line sensing using Raman spectroscopy. The values ​​measured in this way are input into a calculation device or calculation unit 12 to calculate the concentration. The concentration is then controlled by appropriately operating the addition pump and the culture medium CM extraction pump according to the concentration.

[0048] Waste Product Concentration: The waste product concentration is controlled by measuring the waste product concentration in the culture medium CM and adding a medium containing a substance equivalent to the waste product (i.e., a substance that is the basis of the waste product) to achieve a set value based on that value. Examples of waste products include lactic acid and ammonia. The waste product concentration can be measured by sampling the culture medium CM and using HPLC or LC-MS, or by in-line sensing using Raman spectroscopy. The values ​​measured in this way are input into a calculation device or calculation unit 12 to calculate the concentration, and the concentration is controlled by appropriately operating the addition pump and the culture medium CM extraction pump depending on the concentration.

[0049] Viable cell concentration: The viable cell concentration is controlled by measuring the viable cell concentration in the culture medium CM and controlling it to a set value based on that value. If the viable cell concentration is higher than the set value, the bleeding mechanism 36 extracts the culture medium CM containing the viable cells, and the feed mechanism 34 adds an equal amount of medium to lower the concentration. If the viable cell concentration is lower than the set value, the cells are cultured until they grow to the set viable cell count. The viable cell count can be measured by sampling the culture medium CM and counting the cells using a microscope or an image, or by counting inline using capacitance or turbidity.

[0050] [3] Analysis: Quantitative analysis is performed on the objective variable to be maximized or minimized. The objective variable corresponds to, for example, the productivity of a useful substance (concentration, yield relative to input cost, etc.) or quality (activity and molecular structure of the useful substance, etc.). Productivity is quantitatively evaluated by sampling the culture solution CM in the culture tank 30 and measuring the concentrations of components in the culture solution CM. Similarly, quality is quantitatively evaluated by sampling the culture solution CM and analyzing it using HPLC, LC-MS, or the like. In addition to the above-mentioned sampling analysis, the inside of the culture tank 30 can also be measured using an in-line sensor using pH, DO concentration, turbidity, capacitance, etc. Furthermore, the cell state can be measured and analyzed by measuring the exhaust gas and input gas from the culture tank 30, and these measurement results can also be used as the objective variable.

[0051] [4] Input to the calculation unit (including recording) The results of "[1] Setting of experimental conditions" and "[3] Analysis" are input to the calculation unit 12 in preparation for the calculation of "[5] Calculation for proposing the next experimental conditions." When inputting these results to the calculation unit 12, these results can be recorded and used as learning data.

[0052] [5] As an optimization method using computational machine learning to propose the next experimental conditions, Bayesian optimization can be used when there is one objective variable, and multi-objective optimization can be used when there are multiple objective variables. The above methods build a model based on training data and present explanatory variables that result in a higher (or lower) state of the objective variable. It is desirable to display or record the estimated values ​​obtained by the model at this time.

[0053] An example of analysis using multi-objective optimization is described below. Figures 9A to 9D are graphs of functions used in test calculations in multi-objective optimization. Figures 9A and 9C are three-dimensional representations of the graphs. Figure 9B is a two-dimensional representation of the graph in Figure 9A. Figure 9D is a two-dimensional representation of the graph in Figure 9C. Figure 10 is a graph showing an example of a Pareto front for objective variable M (e.g., productivity) and objective variable N (e.g., quality). The Pareto front refers to a line, or in higher dimensions, a surface, formed by a set of Pareto-optimal solutions. In Figure 10, the line connecting the optimal solutions (circles) (line not shown) represents the Pareto front.

[0054] Assume that there are two objective variables (productivity and quality), and that these have the relationship shown in Figures 9A-D depending on the settings of the first and second explanatory variables of the culture conditions. In other words, we assume that productivity and quality fluctuate depending on the culture conditions, with some conditions achieving maximum values ​​or values ​​above standard, while other conditions achieving values ​​below standard. In this regard, by performing multi-objective optimization, a Pareto front indicated by the optimal solution (○), as shown in Figure 10, can be obtained with a small number of trial experiments. Using this Pareto front, users can select culture conditions that maximize productivity and quality while balancing each other.

[0055] [6] Presentation of next experimental conditions The results calculated in "[5] Calculation for proposing next experimental conditions" are displayed or recorded in "[7] Termination condition determination."

[0056] [7] Termination Condition Judgment The termination judgment is terminated when the value of the objective variable preset by the user is satisfied. If not, the process returns to [1] and the next cycle is started with "[6] Presentation of next experimental conditions" as the new input value. For example, if the user presets the objective variable M (e.g., productivity) to be equal to or greater than x and the objective variable N (impurity concentration) to be equal to or less than y, the process ends when the objective variable values ​​in the "[3] Analysis" results are objective variable M≧x and objective variable N≦y. If objective variable M<x or objective variable N>y, the process returns to [1].

[0057] Alternatively, the user does not have to specify the value of the objective variable at the end of the analysis; instead, the user can specify either the maximum or minimum value for the objective variable. In this case, if the objective variable values ​​in the "(3) Analysis" results for the Nth and (N-1)th cycles are deemed equivalent, the process can be considered complete. The user can specify the percentage difference that constitutes equivalence. Furthermore, the equivalence determination (the percentage difference) can be set to a value that is not considered significant due to analytical or experimental variability. The end criterion can also be when the culture conditions in "(1) Experimental Condition Setting" and "(6) Presentation of Next Experimental Conditions" remain unchanged. Furthermore, the end can also be determined when the objective variable estimated from the model in "(5) Calculation for Proposing Next Experimental Conditions" matches the objective variable value in the "(3) Analysis" results within a range that is not significantly different. While the above explanation uses an example with two objective variables, the process can be similarly applied to cases with one or three or more objective variables.

[0058] In the culture system 1 described above, the control unit 13 preferably determines the second culture vessel 30b to which the cells to be cultured are transferred based on the type and value of the second explanatory variable included in the additional experimental information. In this way, the culture experiment can be performed using the second culture vessel 30b that matches the type and value of the second explanatory variable included in the additional experimental information, thereby enabling the culture conditions to be suitably optimized.

[0059] Furthermore, in the culture system 1 described above, when the explanatory variable to be changed in the culture experiment performed in the first culture tank 30a among the plurality of culture tanks 30 in the additional experiment information is the first explanatory variable, the control unit 13 preferably performs control to change the first explanatory variable in the culture experiment performed in the first culture tank 30a. In this case, the culture experiment is performed only in the first culture tank 30a, and the culture conditions can be optimized.

[0060] As described above, the culture system 1 allows for the culture to be continued by changing the culture tank 30 during experimental optimization using machine learning, thereby enabling the optimization of culture conditions, including scale-up, in a short period of time and with a small number of culture experiments. As a result, the culture system 1 can reduce the time and cost required for culture process development.

[0061] <Culture Method> Next, a culture method according to one embodiment of the present invention (hereinafter, sometimes simply referred to as a "main culture method") will be described with reference to Fig. 11. Fig. 11 is a flow chart illustrating the contents of the main culture method.

[0062] This culture method is an experiment for searching for culture conditions for cell culture. As shown in Fig. 11, this culture method includes a calculation step S111 and a control step S112. The calculation step S111 corresponds to the above-mentioned calculation unit 12. The control step S112 corresponds to the above-mentioned control unit 13.

[0063] In the calculation step S111, values ​​of multiple explanatory variables are searched for such that the objective variable in the culture experiment satisfies a predetermined condition based on experimental data from a culture experiment conducted by changing multiple explanatory variables related to the culture conditions. In addition, in the calculation step S111, additional experimental information is output, which is information on the explanatory variables to be changed in the culture experiment based on the searched values ​​of the multiple explanatory variables. As described above, the multiple explanatory variables include one or more first explanatory variables whose values ​​can be changed within a predetermined range in each of the multiple culture tanks 30. Furthermore, the multiple explanatory variables include one or more second explanatory variables whose possible values ​​differ between at least one pair of culture tanks 30 included in the multiple culture tanks 30. In the calculation step S111, when searching for values ​​of the multiple explanatory variables, a machine learning optimization method such as Bayesian optimization can be used when there is one objective variable, or multi-objective optimization can be used when there are multiple objective variables.

[0064] In control step S112, the culture conditions in the multiple culture tanks 30 for the culture experiment are controlled based on the additional experiment information output in calculation step S111, and the transfer of cells to be cultured between the multiple culture tanks 30 is controlled in accordance with a determination described below. That is, in this control step S112, if the explanatory variable to be changed in the culture experiment to be performed in the first culture tank 30a of the multiple culture tanks 30 in the additional experiment information is the second explanatory variable described above, a determination is made as to whether or not the cells to be cultured need to be transferred to a second culture tank 30b different from the first culture tank 30a of the multiple culture tanks 30. Then, in control step S112, if the result indicates that the transfer of cells to be cultured is necessary, the transfer of cells to be cultured from the first culture tank 30a to the second culture tank 30b is controlled.

[0065] According to this culture method, the culture vessel 30 can be changed midway through experimental optimization using machine learning, allowing for continued culture, and therefore, culture conditions, including scale-up, can be optimized in a short period of time with a small number of culture experiments. This allows the culture method to reduce the time and cost required for culture process development.

[0066] <Culture Management Device> Next, a culture management device according to one embodiment of the present invention will be described with reference to Fig. 12. Fig. 12 is an explanatory diagram illustrating an overview of the culture management device 120. Note that the culture management device 120 has a configuration similar to that of the computer 10 of the above-described culture system 1, but differs in that it does not control the culture conditions in the multiple culture tanks 30.

[0067] 12 searches for culture conditions for cell culture. The culture management device 120 includes a calculation unit 121 and a control unit 131. As described above, the calculation unit 121 and the control unit 131 in the culture management device 120 are also configured with, for example, a CPU 11 that executes a program that controls the operation of each component, function, and processing unit.

[0068] That is, as shown in FIG. 12 , the culture management device 120 includes a CPU 11, a RAM 14, a ROM 15, a HDD 16, a communication I / F 17, an input / output I / F 18, and a media I / F 19. The communication I / F 17 is connected to a communication circuit 20. The input / output I / F 18 is connected to an input / output device 21. The media I / F 19 reads and writes data from a recording medium 22. The ROM 15 stores application programs executed by the CPU 11, various data, and the like. The HDD 16 stores experimental data, various data, control programs, and the like as appropriate. The CPU 11 realizes various functions by executing application programs loaded into the RAM 14. The culture management device 120 may be configured as a single computer or may be configured as multiple computers connected via the communication circuit 20.

[0069] The calculation unit 121 searches for values ​​of multiple explanatory variables that will cause the objective variable in the culture experiment to satisfy a predetermined condition, based on experimental data from a culture experiment conducted by changing multiple explanatory variables related to the culture conditions. The calculation unit 121 also outputs additional experimental information, which is information on the explanatory variables to be changed in the culture experiment, based on the searched values ​​of the multiple explanatory variables. As described above, the multiple explanatory variables include one or more first explanatory variables whose values ​​can be changed within a predetermined range in each of the multiple culture tanks 30. The multiple explanatory variables also include one or more second explanatory variables whose possible values ​​differ between at least one pair of culture tanks 30 included in the multiple culture tanks 30. When searching for values ​​of the multiple explanatory variables, the calculation unit 121 can use a machine learning optimization method, such as Bayesian optimization, when there is one objective variable, or multi-objective optimization, when there are multiple objective variables.

[0070] The control unit 131 controls the transfer of cells to be cultured for a culture experiment between the plurality of culture tanks 30. That is, the control unit 131 determines whether or not it is necessary to transfer the cells to be cultured to a second culture tank 30b, which is different from the first culture tank 30a, among the plurality of culture tanks 30, based on at least whether or not the explanatory variable to be changed in the culture experiment to be performed in the first culture tank 30a among the plurality of culture tanks 30 in the additional experiment information is the second explanatory variable.

[0071] The culture management device 120 allows for the culture to be continued by changing the culture tank 30 midway through experimental optimization using machine learning, making it possible to optimize culture conditions, including scale-up, in a short period of time and with a small number of culture experiments. As a result, the culture management device 120 can reduce the time and cost required for culture process development.

[0072] [Optimization of culture conditions using CHO cells] Using the above-mentioned culture-related process optimization system (culture system 1), the culture conditions were optimized for a CHO cell line that produces tissue plasminogen activator (tPA), a glycoprotein. Details are described below.

[0073] (Selection of Objective Variable and Explanatory Variable) Temperature, pH, and DO concentration were selected as explanatory variables (first explanatory variables) that can be set to a wide range of values. Shear stress was selected as an explanatory variable (second explanatory variable) that has a narrower settable range. In descending order of shear stress, a fermenter 30 equipped with a disk turbine blade, a fermenter 30 equipped with a flat paddle, and a fermenter 30 equipped with a pitched paddle were prepared. In addition, antibody concentration was selected as the objective variable. Optimization of the culture conditions was performed using multi-objective optimization.

[0074] (Cells and Medium) For the culture experiments, tPA-producing Chinese hamster ovary cells (CHO cells; CRL-9606 cells) (compatible with both adherent and suspension culture) were purchased from the American Type Culture Collection (ATCC) and used. The medium used was Ham's F12 basal medium supplemented with fetal bovine serum (FBS) (final concentration 10%) and the antibiotics penicillin and streptomycin.

[0075] (Preparation of Suspension Cells (Adaptation from Adherent Cells to Suspension Cells)) CHO cells cultured in an adherent state in a culture flask were detached with trypsin, and the trypsin solution was removed by centrifugation (room temperature, 500 × g, 5 minutes). Then, the cells were diluted with Ham's F12 medium and 1 × 10 cells were collected in a spinner flask. ５ The cells were cultured at a cell concentration of 100 cells / ml in a 100 mL incubator (37°C, 5% CO ２ The cells were cultured under stirring at a temperature of 100°C (temperature 90% or higher, humidity 90% or higher). In this experiment, the cells used were those that had been adapted from adherent cells to suspension cells using the above procedure.

[0076] (Cultivation Operation) The culture conditions were optimized using a culture system 1 equipped with three 1 L-volume culture tanks 30. Aseptic sampling was carried out for each cycle, and the components of the culture solution CM were analyzed.

[0077] (Analysis of Culture Solution Components) The sampled culture solution CM was used to (1) count the number of viable cells, (2) analyze the medium components (glucose, glutamine, lactic acid, ammonia), and (3) quantify tPA protein. The analytical methods for these analyses are described below.

[0078] (1) Counting of viable cells The number of viable cells was counted using a viable cell detector, Vi-CELL (Beckman Coulter). The sampled culture medium CM was placed in the Vi-CELL, and live and dead cells were distinguished by trypan blue staining. Image data of the cells was acquired, and the number of viable cells was obtained by automatic counting.

[0079] (2) Analysis of Medium Components (Glucose, Glutamine, Lactic Acid, Ammonia) Glucose, glutamine, lactic acid, and ammonia in the culture medium CM were measured using Bioplofile 100plus (Nova).

[0080] (3) Quantification of tPA protein: tPA, Human, ELISA Kit (Funakoshi) was used for quantification of tPA protein. The analytical method was based on enzyme-linked immunosorbent assay (ELISA), using mouse anti-tPA antibody as the capture antibody and biotinylated polyclonal anti-tPA antibody as the detection antibody. The substrate was tetramethylbenzidine, and the color reaction was carried out with streptavidin-peroxidase. The absorbance at a wavelength of 450 nm was measured using an absorption spectrophotometer.

[0081] (Optimization using a culture-related process optimization system) To maximize the antibody concentration, the culture condition optimization cycle (the aforementioned cycles [1] to [7]) was repeated using the above culture method. The culture conditions before optimization (culture conditions for the first cycle) were a temperature of 37°C, pH 7.2, dissolved oxygen of 40%, and a pitched paddle stirring impeller (shear force). As the cycle was repeated, the culture vessel 30 used was changed depending on the range of shear stress, and the culture medium CM was also fed accordingly.

[0082] The difference between the antibody concentration at the 20th cycle and that at the 9th cycle was 5% or less, so the culture was terminated at the 20th cycle. The culture conditions at the 20th cycle (values ​​of the first explanatory variable and the second explanatory variable) were set as the optimal culture conditions (temperature 36°C, pH 7.0, dissolved oxygen 20%, stirring blade (shear force) disk turbine blade).

[0083] The results of the culture after optimization of the culture conditions are shown in Figure 13. Figure 13 is a graph showing the relationship between the number of culture days (days) and the antibody concentration under the conditions before optimization (conditions for the first cycle) and after optimization (culture conditions for the 20th cycle). As shown in Figure 13, an improvement in the antibody concentration was confirmed under the optimized conditions compared to the conditions before optimization. The optimization described above uses four explanatory variables. Conventional methods that perform optimization at five levels require the experimental investigation of 4^5 = 1,024 experimental conditions. However, optimization using the culture system 1 allowed the optimal conditions to be found after just 20 experiments. The cost of consumables required for the experiment was reduced to 1 / 20 of that required for the conventional method. Furthermore, the experimental period was reduced to 1 / 6 or less of that required for the conventional method using eight consecutive culture vessels. It was confirmed that the present invention can reduce the time and cost required for culture process development.

[0084] REFERENCE SIGNS LIST 1 Cultivation system 12, 121 Calculation unit 13, 131 Control unit 30 Cultivation tank 30a First cultivator 30b Second cultivator 31 Gas supply valve 32a Transfer valve 32b Transfer valve 33 Gas exhaust valve 34 Feed mechanism 35 Cell separation mechanism 36 Bleeding mechanism 120 Cultivation management device

Claims

1. A culture system for conducting experiments to search for culture conditions in cell culture, comprising: a calculation unit that searches for values ​​of a plurality of explanatory variables that will cause a target variable in a culture experiment to satisfy a predetermined condition based on experimental data from a culture experiment conducted by changing a plurality of explanatory variables related to the culture conditions, and outputs additional experimental information that is information on the explanatory variables to be changed in the culture experiment based on the searched values ​​of the plurality of explanatory variables; a plurality of culture vessels for the culture experiment; and a control unit that controls the culture conditions in the plurality of culture vessels based on the additional experimental information output by the calculation unit and controls the transfer of cells to be cultured between the plurality of culture vessels in accordance with a determination, wherein the plurality of explanatory variables include one or more first explanatory variables whose values ​​can be changed within a predetermined range in each of the plurality of culture vessels, and one or more second explanatory variables whose possible values ​​differ between at least one pair of culture vessels included in the plurality of culture vessels, When the explanatory variable to be changed in the culture experiment performed in a first culture tank among the plurality of culture tanks in the additional experimental information is the second explanatory variable, the control unit determines whether or not it is necessary to transfer the cells to be cultured to a second culture tank among the plurality of culture tanks that is different from the first culture tank, and when it is determined that the transfer is necessary, controls the transfer of the cells to be cultured from the first culture tank to the second culture tank.

2. The culture system described in claim 1, characterized in that the control unit determines the second culture tank to which the cells to be cultured are transferred based on the type and value of the second explanatory variable included in the additional experimental information.

3. The culture system described in claim 1, characterized in that the control unit controls to change the first explanatory variable in the culture experiment conducted in the first culture tank when the explanatory variable to be changed in the culture experiment conducted in a first culture tank among the multiple culture tanks in the additional experimental information is the first explanatory variable.

4. The culture system according to claim 1, wherein the plurality of culture tanks are equipped with a gas supply valve that supplies gas to the first culture tank, a transfer valve that connects the first culture tank and the second culture tank, and a gas exhaust valve that exhausts gas from the second culture tank, and the control unit, when transferring the cells to be cultured from the first culture tank to the second culture tank, opens the gas supply valve to start supplying gas to the first culture tank, and then opens the transfer valve and the gas exhaust valve to transfer the cells to be cultured.

5. The culture system according to claim 1, wherein the culture tank comprises a feed mechanism, a cell separation mechanism, a bleeding mechanism, and mechanisms for controlling temperature, pH, and dissolved oxygen.

6. The culture system according to claim 1, wherein the second explanatory variable is at least one of shear stress and bubble diameter.

7. The culture system according to claim 6, wherein the stirring blade for controlling the shear stress is at least one of a pitched paddle, a flat paddle, and a disk turbine blade.

8. A culture method for conducting an experiment to search for culture conditions in cell culture, comprising: a calculation step of searching for values ​​of a plurality of explanatory variables that will cause a target variable in the culture experiment to satisfy a predetermined condition based on experimental data from a culture experiment conducted by changing a plurality of explanatory variables related to the culture conditions, and outputting additional experimental information that is information on the explanatory variables to be changed in the culture experiment based on the searched values ​​of the plurality of explanatory variables; and a control step of controlling the culture conditions in a plurality of culture vessels for the culture experiment based on the additional experimental information output in the calculation step, and controlling the transfer of cells to be cultured between the plurality of culture vessels in accordance with determination, wherein the plurality of explanatory variables include one or more first explanatory variables whose values ​​can be changed within a predetermined range in each of the plurality of culture vessels, and one or more second explanatory variables whose possible values ​​differ between at least one pair of culture vessels included in the plurality of culture vessels, The control step determines whether or not the cells to be cultured need to be transferred to a second culture tank among the plurality of culture tanks that is different from the first culture tank when the explanatory variable to be changed in the culture experiment performed in a first culture tank among the plurality of culture tanks in the additional experimental information is the second explanatory variable, and controls the transfer of the cells to be cultured from the first culture tank to the second culture tank when it is determined that the transfer is necessary.

9. A culture management device for searching for culture conditions in cell culture, comprising: a calculation unit that searches for values ​​of a plurality of explanatory variables that will cause a target variable in a culture experiment to satisfy a predetermined condition based on experimental data from a culture experiment conducted by changing a plurality of explanatory variables related to the culture conditions, and outputs additional experimental information that is information on the explanatory variables to be changed in the culture experiment based on the searched values ​​of the plurality of explanatory variables; and a control unit that controls the transfer of cells to be cultured between a plurality of culture tanks for the culture experiment, wherein the plurality of explanatory variables include one or more first explanatory variables whose values ​​can be changed within a predetermined range in each of the plurality of culture tanks, and one or more second explanatory variables whose possible values ​​differ between at least one pair of culture tanks included in the plurality of culture tanks, and the control unit determines whether or not it is necessary to transfer the cells to be cultured to a second culture tank different from the first culture tank among the plurality of culture tanks, based at least on whether the explanatory variable to be changed in the culture experiment conducted in a first culture tank among the plurality of culture tanks in the additional experimental information is the second explanatory variable.

Images