Method, device, equipment and medium for generating culture medium formula

Abstract

The present invention provides a method, device, equipment and medium for generating culture medium formula, which can generate culture medium formula data with uniform composition, reduce subjective interference caused by human factors in the experimental design process, lower the formula design threshold, labor cost and time cost, shorten the liquid preparation time, thereby improving work efficiency and further improving the learning effect of the machine learning model.

Description

Technical Field

[0001] The present invention relates to the field of biotechnology, and in particular to a method, device, equipment and medium for generating a culture medium formula. Background Art

[0002] Culture medium refers to a nutrient matrix composed of a combination of different nutrients that supports the growth and reproduction of microorganisms, plants, or animals. It generally contains several major categories of substances, such as carbohydrates, nitrogen-containing substances, inorganic salts, vitamins, and water. Culture medium is not only the basic substance for providing cell nutrition and promoting cell proliferation, but also the living environment for cell growth and reproduction. The traditional method of developing culture medium formula is to use one or several classic culture media as the basis, add a variety of different ingredients, and use single-factor experiments or DOE experiments to find the key components. Then, use multiple DOE experimental designs such as response surface curves to optimize the concentration of each component to obtain the optimal formula; or optimize the formula based on cell metabolism analysis, genomic analysis, and proteomic analysis to find the changes of each component during cell growth and their impact on the yield and quality of the target product.

[0003] However, the present inventors found in their research on the prior art that the traditional culture medium formulation development method can easily lead to multicollinearity in the formulation experimental data. Summary of the Invention

[0004] The present invention provides a culture medium formula generation method, device, equipment and medium, which can reduce the probability of multicollinearity problems occurring during the formula design experiment process.

[0005] To achieve the above objectives, in a first aspect, an embodiment of the present invention provides a method for generating a culture medium formula, comprising the following steps:

[0006] Obtaining a trained hypersphere model; wherein each hypersphere coordinate on the hypersphere model corresponds to each mother liquid formula, and each mother liquid formula includes a plurality of components;

[0007] Obtaining a target concentration range for each component in the culture medium that meets the target formulation requirements, and obtaining target parameter information corresponding to each component based on the target concentration range of each component;

[0008] According to the target parameter information corresponding to each component, the concentration range of each component corresponding to the hypersphere model is adjusted to obtain a culture medium formula that meets the target formula requirements.

[0009] As one embodiment of the first aspect, obtaining a trained hypersphere model specifically includes:

[0010] Obtaining the hyperspherical coordinates of each mother solution formula in the culture medium formula sample on the hypersphere;

[0011] Construct a loss function based on the hyperspherical coordinates of each mother solution formula on the hypersphere;

[0012] According to the loss function, the gradient descent algorithm is used to optimize the parameters of the hypersphere model to obtain the trained hypersphere model.

[0013] As one embodiment of the first aspect, the target parameter information corresponding to each component includes: the median concentration value of each component within the target concentration range and the multiple of the concentration difference between the median concentration value of each component and the upper limit or lower limit of the target concentration range;

[0014] Then, according to the target parameter information corresponding to each component, the concentration range of each component corresponding to the hypersphere model is adjusted to obtain a culture medium formula that meets the target formula requirements, specifically including:

[0015] Obtain the hyperspherical coordinates of each mother liquid formula on the hyperspherical model;

[0016] The hyperspherical coordinates of each mother liquid formula on the hypersphere are converted into coordinates in the Cartesian coordinate system to obtain the x value of each component in each mother liquid formula in the Cartesian coordinate system;

[0017] For each component of each stock solution formula in the culture medium, the sum of the product of the x value of the component in the Cartesian coordinate system and the multiple of the concentration difference of the component plus the median concentration value of the component is used as the target concentration value of the component;

[0018] According to the target concentration value of each component of each mother solution formula in the culture medium, a culture medium formula that meets the target formula requirements is obtained.

[0019] As one embodiment of the first aspect, for each component of each stock solution formula in the culture medium, the target concentration value of the component is calculated as the sum of the product of the x value of the component in the Cartesian coordinate system and the multiple of the concentration difference of the component plus the median concentration value of the component:

[0020] Y 母液成分值 =X 超球面 *d+Center

[0021] Among them, Y 母液成分值 Indicates a component of the culture medium formula that meets the target formulation requirements, X 超球面 It represents the x value of the component in the Cartesian coordinate system, d represents the multiple of the concentration difference of the component, and Center represents the middle concentration value of the component.

[0022] As one embodiment of the first aspect, the method for generating a culture medium formula further includes:

[0023] The culture medium formulas that meet the target formula requirements are divided into multiple formula groups, and each formula group is mixed according to the proportion to generate a mixed culture medium formula group.

[0024] As one embodiment of the first aspect, the culture medium formula that meets the target formula requirements is divided into multiple formula groups, and each formula group is mixed according to a proportion to generate a mixed culture medium formula group, specifically including:

[0025] Divide the culture medium formula into culture medium formula groups with preset equal portions;

[0026] For each culture medium formula group, randomly extracting a preset number of formulas from the culture medium formula group, and mixing the extracted formulas according to a randomly generated ratio to obtain a mixed culture medium formula;

[0027] For each culture medium formula group, the operation of obtaining the mixed culture medium formula is repeatedly performed until the iteration number threshold is met, and a mixed culture medium formula group is generated.

[0028] As one embodiment of the first aspect, for each culture medium formula group, a preset number of formulas are randomly selected from the culture medium formula group, and the selected formulas are mixed according to a randomly generated ratio. The calculation formula for the mixed culture medium formula is:

[0029]

[0030] Where Y represents the mixed culture medium formula, c′ represents a random number in the range of 0 to 100 generated by uniform distribution, X i Represents a randomly selected recipe i.

[0031] In a second aspect, an embodiment of the present invention provides a culture medium formula generating device, comprising:

[0032] A hypersphere model acquisition module is used to acquire a trained hypersphere model, wherein each hypersphere coordinate on the hypersphere model corresponds to each mother liquid formula, and each mother liquid formula includes a number of components;

[0033] The formula parameter information acquisition module is used to obtain the target concentration range of each component in the culture medium that meets the target formula requirements, and obtain the target parameter information corresponding to each component based on the target concentration range of each component;

[0034] The target culture medium formula generation module is used to adjust the concentration range of each component corresponding to the hypersphere model according to the target parameter information corresponding to each component, so as to obtain a culture medium formula that meets the target formula requirements.

[0035] In a third aspect, an embodiment of the present invention provides a terminal device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the method for generating a culture medium formula of any embodiment of the first aspect described above is implemented.

[0036] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the culture medium formula generation method of any embodiment of the first aspect above.

[0037] Compared with the existing technology, the embodiment of the present invention provides a method, device, equipment and medium for generating a culture medium formula. By generating high-quality culture medium formula data with uniform composition on a hypersphere, the probability of multicollinearity problems occurring during the formula design experiment is reduced, and the learning effect of the machine learning model in simulating culture media with different content components to predict target yields during training is further improved, thereby shortening the product development cycle. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 1 is a flow chart of a method for generating a culture medium formula provided by an embodiment of the present invention;

[0039] Figure 2 This is a schematic structural diagram of a culture medium formula generating device provided by an embodiment of the present invention;

[0040] Figure 3 It is a structural diagram of a terminal device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0042] In the first aspect, the present invention provides a method for generating a culture medium formula, see Figure 1 , is a flow chart of a method for generating a culture medium formula provided by an embodiment of the present invention, the method comprising steps S11 to S13:

[0043] S11. Obtaining a trained hypersphere model; wherein each hypersphere coordinate on the hypersphere model corresponds to each mother liquid formula, and each mother liquid formula includes several components.

[0044] It can be understood that a one-dimensional sphere is a circle on a two-dimensional plane, a two-dimensional sphere is an ordinary sphere in three-dimensional space, and a hypersphere is the generalization of an ordinary sphere in any dimension.

[0045] S12. Obtain a target concentration range of each component in the culture medium that meets the target formulation requirements, and obtain target parameter information corresponding to each component based on the target concentration range of each component.

[0046] S13. According to the target parameter information corresponding to each component, the concentration range of each component corresponding to the hypersphere model is adjusted to obtain a culture medium formula that meets the target formula requirements.

[0047] It is understandable that in traditional culture medium formulation experimental design, formulators need to use their extensive professional theoretical knowledge of basic chemistry, biochemistry and molecular biology, cell biology, etc. to conduct experiments based on specific optimization goals. Each experiment takes a long time to prepare the solution, and the design threshold is high, and the labor and time costs are huge. Currently, traditional biotechnology generally uses DOE experiments for analysis in the process of formula experimental design. During the experimental formula design process, the components of the array will be changed proportionally. The experimental formula data formed by this method often leads to multicollinearity in the data, thereby reducing the machine learning model's ability to learn the experimental formula data. In addition, the process of accumulating the sample formula data required for training the machine learning model through experimental design in traditional biotechnology is usually time-consuming.

[0048] Compared with the existing technology, the culture medium formula generation method provided by the embodiment of the present invention can address the multicollinearity problem of experimental formula data that cannot be solved by traditional experimental design ideas. It adopts a multidimensional hypersphere uniform composition culture medium formula design method to eliminate data multicollinearity and improve the efficient learning of data by machine learning models during the training process. Among them, the culture medium formula data with uniform composition can be understood as the sample formula points being evenly distributed on the hypersphere.

[0049] As an optional embodiment of the first aspect, step S11 specifically includes:

[0050] S111. Obtain the hyperspherical coordinates of each mother solution formula in the culture medium formula sample on the hypersphere.

[0051] S112. Construct a loss function according to the hyperspherical coordinates of each mother solution formula on the hypersphere.

[0052] S113. According to the loss function, the parameters of the hypersphere model are optimized using a gradient descent algorithm to obtain a trained hypersphere model.

[0053] For example, a circle in a two-dimensional plane is represented by formula (1), wherein any position in polar coordinates (r, θ) can be represented by an angle θ and a distance r from the origin to the pole.

[0054]

[0055] Among them, x 2 +y 2 =r 2 .

[0056] The spherical surface in three-dimensional space can be expressed by formula (2), where any position in the spherical coordinates (r, α, θ) can be represented by the radial distance r, the azimuth angle α, and the polar angle θ.

[0057]

[0058] Among them, x 2 +y 2 +z 2 =r 2 .

[0059] Similarly, the coordinate system in multidimensional space is similar to the coordinate system in three-dimensional space and is expressed by formula (3).

[0060]

[0061] in, Hyperspherical coordinates (r,θ1,θ2…θ n-1 ) can be obtained by the radial distance r, angle θ1, θ2…θ n-1 To express it, where the value range of the last angle is 2π, and the value range of the other angles is π, which covers the entire sphere.

[0062] In order to make the distance distribution between points on the hypersphere as uniform as possible, the gradient descent algorithm is used to minimize the loss function. The specific calculation formulas are shown in (4) to (7).

[0063] A point on the hypersphere can be represented by either a Cartesian coordinate system or a spherical coordinate system:

[0064]

[0065] The distance formula between points on the hypersphere is:

[0066]

[0067] The formula of the loss function is:

[0068]

[0069] Using gradient descent:

[0070]

[0071] Specifically, in this embodiment, the model is trained according to the above-mentioned derivation formula of the hypersphere in the multidimensional space in combination with the gradient descent method, thereby obtaining a hypersphere model in which the sample formula points are sufficiently evenly distributed in the multidimensional space.

[0072] It should be noted that each point on the hypersphere represents a mother liquor formula, and each mother liquor formula is composed of multiple components. The sample formula points are different from each other and are evenly distributed on the sphere. The spatial distribution distance between adjacent points is equal, thereby obtaining information on the components corresponding to the sample formula points evenly distributed on the sphere in the hypersphere.

[0073] As one optional embodiment of the first aspect, the target parameter information corresponding to each component includes: a median concentration value of each component within the target concentration range and a multiple of a concentration difference between the median concentration value of each component and an upper limit or a lower limit of the target concentration range;

[0074] Then step S13 specifically includes:

[0075] S131. Obtain the hyperspherical coordinates of each mother liquid formula on the hyperspherical model.

[0076] S132. Convert the hyperspherical coordinates of each mother liquid formula on the hypersphere into coordinates in a Cartesian coordinate system, and obtain the x value of each component in each mother liquid formula in the Cartesian coordinate system.

[0077] S133. For each component of each mother solution formula in the culture medium, the sum of the product of the x value of the component in the Cartesian coordinate system and the multiple of the concentration difference of the component plus the median concentration value of the component is used as the target concentration value of the component.

[0078] S134. Obtain a culture medium formula that meets the target formula requirements based on the target concentration value of each component of each mother solution formula in the culture medium.

[0079] Specifically, the component range of the culture medium formula corresponding to the target cell line to be tested is obtained, and for each component, the upper and lower boundary values ​​of its target concentration are determined, and the middle concentration value and the multiple of the concentration difference are determined based on the upper and lower limits.

[0080] It should be noted that the median concentration of each component within the target concentration range can be determined by searching the upper and lower limits of the range, or by the median value corresponding to the optimal formula of historical samples. This can be set based on actual needs or experiments and is not limited here. Similarly, the multiple corresponding to the concentration difference can be single, double, triple, etc., and can be set based on actual needs or experiments and is not limited here.

[0081] As an optional embodiment of the first aspect, the calculation formula of step S133 is:

[0082] Y 母液成分值 =X 超球面 *d+Center (8)

[0083] Among them, Y 母液成分值 Indicates a component of the culture medium formula that meets the target formulation requirements, X 超球面 It represents the x value of the component in the Cartesian coordinate system, d represents the multiple of the concentration difference of the component, and Center represents the middle concentration value of the component.

[0084] It can be understood that each point on the hypersphere represents a mother liquor formula, and each point consists of multiple Y 母液成分值 Specifically, from the above formula (4), the x value of each component in the Cartesian coordinate system is expressed as x1, x2, ..., x n .

[0085] It is worth noting that the culture medium formula formulated by randomly generating uniform components on a multi-dimensional hypersphere effectively reduces the multicollinearity of sample data caused by the traditional biotechnology using DOE experiments for formula design, improves the efficient learning of data by machine learning models during the training process, and is the basis for quickly understanding cell culture theoretical knowledge and effectively mastering the impact of each component on target yield during machine learning model training, providing data support for the analysis results.

[0086] As an optional embodiment of the first aspect, the method for generating a culture medium formula further includes step S14:

[0087] S14. Divide the culture medium formula that meets the target formula requirements into multiple formula groups, and mix each formula group according to the proportion to generate a mixed culture medium formula group.

[0088] As an optional embodiment of the first aspect, S14 specifically includes:

[0089] S141. Divide the culture medium formula into culture medium formula groups of preset equal portions.

[0090] Illustratively, after obtaining several culture medium formulas in step S13, they are divided into M culture medium formula groups, each of which contains m culture medium formulas.

[0091] S142. For each culture medium formula group, randomly extract a preset number of formulas from the culture medium formula group, and mix the extracted formulas according to a randomly generated ratio to obtain a mixed culture medium formula.

[0092] Exemplarily, in each culture medium formula group containing m mother solution formulas, n culture medium formulas X are randomly selected, and the culture medium formulas are mixed according to a randomly generated ratio to obtain a mixed culture medium formula containing m′ mixed formulas.

[0093] S143. For each culture medium formula group, repeatedly perform the operation of obtaining the mixed culture medium formula until the iteration number threshold is met, and a mixed culture medium formula group is generated.

[0094] Exemplarily, the operation of extracting n culture medium formulas from each group of culture medium formulas is repeated until the iteration number threshold K is met. Then, the number of formula groups in the M groups of culture medium formulas after repeated execution is M*m, and the number of mixed culture medium formulas obtained is M*m′*K. Then, the number of culture medium formulas included in the mixed culture medium formula group is M*m+M*m′*K.

[0095] It is worth noting that by mixing culture medium formulas with uniform hyperspherical composition, a large amount of new sample formula data can be quickly generated, simplifying the complexity of the biological experiment formulation process, lowering the threshold for formulation development, and greatly shortening the time period for accumulating sample formula data required for machine learning model training. At the same time, by constructing a high-quality and substantial database of mother solution formulas, and subsequently rapidly proliferating the database by randomly mixing mother solution formulas, the problems of low formulation efficiency caused by the complex composition of existing culture media and the low learning quality of sample formula data for machine learning models are solved. This further improves the formulation efficiency of formulation staff and reduces the possibility of formulation errors caused by human factors.

[0096] As one embodiment of the first aspect, the calculation formula of step S142 is:

[0097]

[0098] Where Y represents the mixed culture medium formula, c′ represents a random number in the range of 0 to 100 generated by uniform distribution, X i Represents a randomly selected recipe i.

[0099] As an optional embodiment of the first aspect, after obtaining the target mother solution formula, the following calculation formula is used to detect whether the target mother solution formula has multicollinearity:

[0100]

[0101] in, represents the i-th independent variable x i The coefficient of determination between it and other independent variables.

[0102] Specifically, taking component A as the dependent variable and components B and C as independent variables, we can get the R of the regression equation. 2 , and the VIF value of component A is obtained using formula (10). Similarly, the VIF values ​​of components B and C can be obtained. When the variance inflation factor exceeds 100, there is severe multicollinearity; when the variance inflation factor is between 10 and 100, there is strong multicollinearity; when the variance inflation factor is less than 10, there is no multicollinearity. The larger the variance inflation factor (VIF), the greater the possibility of multicollinearity between the independent variables.

[0103] It can be understood that multicollinearity refers to the presence of linear correlations between independent variables, meaning that one independent variable is a linear combination of one or more other independent variables. The Variance Inflation Factor (VIF) is often calculated to test whether a regression model has severe multicollinearity issues. High levels of multicollinearity between formulation components can lead to poor predictions of target yields by machine learning models.

[0104] In order to further demonstrate the technical effects achieved by the method for generating a culture medium formula provided by the present invention, the present invention is further described below in conjunction with application examples during the research and development process of the inventors of the present invention:

[0105] Assume that 140 culture medium formulas are obtained according to step S13 above. The 140 culture medium formulas are divided into 7 groups, each of which contains 20 culture medium formulas. Then, for each culture medium formula group, 4 formulas are randomly selected from the group, each of which is called X. The culture medium formulas are mixed in a certain ratio to obtain 20 mixed culture medium formulas.

[0106] Repeat the process of randomly selecting 4 formulas from each culture medium formula group and The mother liquor formula is mixed in the same proportion, and this process is repeated 6 times, then the mixed culture medium formula obtained is 7*(20+6*20), that is, 980 mixed culture medium formulas.

[0107] After obtaining 980 mixed culture medium formulas, we can perform a multicollinearity test on the new formula data formed by the mother solution and the mixed mother solution by calculating the variance inflation factor (VIF). If the variance inflation factor is less than 10, it means that the sample data does not have multicollinearity.

[0108] In a second aspect, an embodiment of the present invention provides a culture medium formula generating device, see Figure 2 , is a schematic structural diagram of a culture medium formula generating device provided by an embodiment of the present invention, comprising:

[0109] A hypersphere model acquisition module 21 is used to acquire a trained hypersphere model, wherein each hypersphere coordinate on the hypersphere model corresponds to each mother liquid formula, and each mother liquid formula includes several components;

[0110] The formula parameter information acquisition module 22 is used to obtain the target concentration range of each component in the culture medium that meets the target formula requirements, and obtain the target parameter information corresponding to each component based on the target concentration range of each component;

[0111] The target culture medium formula generation module 23 is used to adjust the concentration range of each component corresponding to the hypersphere model according to the target parameter information corresponding to each component, so as to obtain a culture medium formula that meets the target formula requirements.

[0112] Compared with the existing technology, the culture medium formula generation device provided by the embodiment of the present invention reduces the probability of multicollinearity problems occurring during the formula design experiment by generating high-quality culture medium formula data with uniform composition on a hypersphere, further improves the learning effect of the machine learning model in simulating culture media with different content components to predict target yield during training, and shortens the product development cycle.

[0113] As one embodiment of the second aspect, the hypersphere model acquisition module 21 is specifically configured to:

[0114] Obtaining the hyperspherical coordinates of each mother solution formula in the culture medium formula sample on the hypersphere;

[0115] Construct a loss function based on the hyperspherical coordinates of each mother solution formula on the hypersphere;

[0116] According to the loss function, the gradient descent algorithm is used to optimize the parameters of the hypersphere model to obtain the trained hypersphere model.

[0117] As one embodiment of the second aspect, the target parameter information corresponding to each component includes: the median concentration value of each component within the target concentration range and the multiple of the concentration difference between the median concentration value of each component and the upper limit or lower limit of the target concentration range;

[0118] Then, the culture medium formula generating module 23 is specifically used for:

[0119] Obtain the hyperspherical coordinates of each mother liquid formula on the hyperspherical model;

[0120] The hyperspherical coordinates of each mother liquid formula on the hypersphere are converted into coordinates in the Cartesian coordinate system to obtain the x value of each component in each mother liquid formula in the Cartesian coordinate system;

[0121] For each component of each stock solution formula in the culture medium, the sum of the product of the x value of the component in the Cartesian coordinate system and the multiple of the concentration difference of the component plus the median concentration value of the component is used as the target concentration value of the component;

[0122] According to the target concentration value of each component of each mother solution formula in the culture medium, a culture medium formula that meets the target formula requirements is obtained.

[0123] As one embodiment of the second aspect, for each component of each stock solution formula in the culture medium, the target concentration value of the component is calculated as the sum of the product of the x value of the component in the Cartesian coordinate system and the multiple of the concentration difference of the component plus the median concentration value of the component:

[0124] Y 母液成分值 =X 超球面 *d+Center

[0125] Among them, Y 母液成分值 Indicates a component of the culture medium formula that meets the target formulation requirements, X 超球面 It represents the x value of the component in the Cartesian coordinate system, d represents the multiple of the concentration difference of the component, and Center represents the middle concentration value of the component.

[0126] As one embodiment of the second aspect, the culture medium formula generating device further includes:

[0127] The mixed culture medium formula generation module 24 is used to divide the culture medium formulas that meet the target formula requirements into multiple formula groups, and mix each formula group according to the proportion to generate a mixed culture medium formula.

[0128] As one embodiment of the second aspect, the mixed culture medium formula generation module 24 is specifically configured to:

[0129] Divide the culture medium formula into culture medium formula groups with preset equal portions;

[0130] For each culture medium formula group, randomly extracting a preset number of formulas from the culture medium formula group, and mixing the extracted formulas according to a randomly generated ratio to obtain a mixed culture medium formula;

[0131] For each culture medium formula group, the operation of obtaining the mixed culture medium formula is repeatedly performed until the iteration number threshold is met, and a mixed culture medium formula group is generated.

[0132] As one embodiment of the second aspect, for each culture medium formula group, a preset number of formulas are randomly selected from the culture medium formula group, and the selected formulas are mixed according to a randomly generated ratio. The calculation formula for the mixed culture medium formula is:

[0133]

[0134] Where Y represents the mixed culture medium formula, c′ represents a random number in the range of 0 to 100 generated by uniform distribution, X i Represents a randomly selected recipe i.

[0135] As an optional embodiment of the second aspect, after obtaining the target mother solution formula, the following calculation formula is used to detect whether the target mother solution formula has multicollinearity:

[0136]

[0137] in, represents the i-th independent variable x i The coefficient of determination between it and other independent variables.

[0138] In addition, it should be noted that the specific implementation schemes and beneficial effects of each embodiment of a culture medium formula generation device provided in the second aspect of an embodiment of the present invention are the same as the specific implementation schemes and beneficial effects of each embodiment of a culture medium formula generation method provided in the first aspect of an embodiment of the present invention, and will not be repeated here.

[0139] In a third aspect, an embodiment of the present invention provides a terminal device, see Figure 3 , is a schematic diagram of the structure of a terminal device provided in an embodiment of the present invention. The terminal device 3 in this embodiment includes: a processor 30, a memory 31, and a computer program stored in the memory 31 and executable by the processor 30. When the processor 30 executes the computer program, it implements the culture medium formula generation method described in any of the embodiments of the first aspect. Alternatively, when the processor 30 executes the computer program, it implements the functions of each module described in the aforementioned device embodiments.

[0140] Exemplarily, the computer program may be divided into one or more modules, one or more of which are stored in the memory 31 and executed by the processor 30 to implement the present invention. One or more modules may be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program in the terminal device 3.

[0141] Terminal device 3 can be a computing device such as a desktop computer, laptop, PDA, or cloud server. Terminal device 3 may include, but is not limited to, a processor 30 and a memory 31. Those skilled in the art will appreciate that the schematic diagram is merely an example of a terminal device and does not limit the terminal device. The terminal device may include more or fewer components than shown, or a combination of certain components, or different components. For example, terminal device 3 may also include input / output devices, network access devices, buses, and the like.

[0142] The processor 30 may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor. The processor 30 is the control center of the terminal device 3 and connects various parts of the entire terminal device 3 using various interfaces and lines.

[0143] Memory 31 can be used to store computer programs and / or modules. Processor 30 implements various functions of terminal device 3 by running or executing computer programs and / or modules stored in memory 31 and accessing data stored in memory 31. Memory 31 may primarily include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function (such as a sound playback function or an image playback function); the data storage area may store data generated based on the use of the mobile phone (such as audio data, a phone book, etc.). In addition, memory 31 may include high-speed random access memory and non-volatile memory, such as a hard disk, internal memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0144] Among them, if the module integrated in the terminal device 3 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor 30, it can implement the steps of the above-mentioned various method embodiments. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0145] It should be noted that the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement the present invention without inventive effort.

[0146] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the culture medium formula generation method as described above.

[0147] Those skilled in the art will appreciate that the modules in the devices in the embodiments can be adaptively changed and set in one or more devices different from the embodiments. The modules or units in the embodiments can be combined into one module or unit, and in addition they can be divided into multiple submodules or subunits. Except that at least some of such features and / or processes or units are mutually exclusive, all features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed in this manner can be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) can be replaced by an alternative feature that provides the same, equivalent or similar purpose.

[0148] It should be noted that the above embodiments illustrate rather than limit the present invention, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements.

[0149] Furthermore, those skilled in the art will appreciate that although some embodiments herein include certain features included in other embodiments but not other features, combinations of features from different embodiments are intended to be within the scope of the present invention and to form different embodiments. For example, in the claims of the present invention, any of the claimed embodiments may be used in any combination.

[0150] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for generating a culture medium formula, characterized in that: include: Obtaining a trained hypersphere model; wherein each hypersphere coordinate on the hypersphere model corresponds to each mother liquid formula, and each mother liquid formula includes a plurality of components; Obtaining a target concentration range for each component in the culture medium that meets the target formulation requirements, and obtaining target parameter information corresponding to each component based on the target concentration range of each component; According to the target parameter information corresponding to each of the components, the concentration range of each component corresponding to the hypersphere model is adjusted to obtain a culture medium formula that meets the target formula requirements.

2. The method for generating a culture medium formula according to claim 1, wherein The step of obtaining the trained hypersphere model specifically includes: Obtaining the hyperspherical coordinates of each mother solution formula in the culture medium formula sample on the hypersphere; Constructing a loss function according to the hyperspherical coordinates of each of the mother solution formulas on the hypersphere; According to the loss function, the parameters of the hypersphere model are optimized using a gradient descent algorithm to obtain a trained hypersphere model.

3. The method for generating a culture medium formula according to claim 1, wherein: The target parameter information corresponding to each component includes: the median concentration value of each component within the target concentration range and the multiple of the concentration difference between the median concentration value of each component and the upper limit or lower limit of the target concentration range; Then, according to the target parameter information corresponding to each component, the concentration range of each component corresponding to the hypersphere model is adjusted to obtain a culture medium formula that meets the target formula requirements, specifically including: Obtain the hyperspherical coordinates of each mother liquid formula on the hyperspherical model; Converting the hyperspherical coordinates of each of the mother solution formulas on the hypersphere into coordinates in a Cartesian coordinate system to obtain the x value of each component in each of the mother solution formulas in the Cartesian coordinate system; For each component of each stock solution formulation in the culture medium, the target concentration value of the component is calculated using the following formula: Y 母液成分值 =X 超球面 *d+Center Among them, Y 母液成分值 Indicates a component of the culture medium formula that meets the target formulation requirements, X 超球面 represents the x value of the component in the Cartesian coordinate system, d represents the multiple of the concentration difference of the component, and Center represents the middle concentration value of the component; According to the target concentration value of each component of each mother solution formula in the culture medium, a culture medium formula that meets the target formula requirements is obtained.

4. The method for generating a culture medium formula according to claim 1, wherein: The method further comprises: The culture medium formula that meets the target formula requirements is divided into multiple formula groups, and each formula group is mixed according to a proportion to generate a mixed culture medium formula group.

5. The method for generating a culture medium formula according to claim 4, wherein: The method of dividing the culture medium formula that meets the target formula requirements into multiple formula groups, and mixing each formula group according to a proportion to generate a mixed culture medium formula group specifically includes: Dividing the culture medium formula into culture medium formula groups of predetermined equal portions; For each of the culture medium formula groups, randomly extracting a preset number of formulas from the culture medium formula group, and mixing the extracted formulas according to a randomly generated ratio to obtain a mixed culture medium formula; For each culture medium formula group, the operation of obtaining the mixed culture medium formula is repeatedly performed until a threshold number of iterations is met, and a mixed culture medium formula group is generated.

6. The method for generating a culture medium formula according to claim 5, wherein: For each of the culture medium formula groups, a preset number of formulas are randomly extracted from the culture medium formula group, and the extracted formulas are mixed according to a randomly generated ratio to obtain the mixed culture medium formula. The calculation formula is: Wherein, Y represents the medium formula after mixing, c ′ Indicates a random number in the range of 0 to 100 generated by a uniform distribution. X i Represents a randomly selected recipe i.

7. A culture medium formula generating device, characterized in that: include: A hypersphere model acquisition module is used to acquire a trained hypersphere model; wherein each hypersphere coordinate on the hypersphere model corresponds to each mother liquid formula, and each mother liquid formula includes a plurality of components; A formula parameter information acquisition module is used to obtain the target concentration range of each component in the culture medium that meets the target formula requirements, and obtain the target parameter information corresponding to each component according to the target concentration range of each component; The target culture medium formula generation module is used to adjust the concentration range of each component corresponding to the hypersphere model according to the target parameter information corresponding to each component, so as to obtain a culture medium formula that meets the target formula requirements.

8. A terminal device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, the method for generating a culture medium formula according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the culture medium formula generation method according to any one of claims 1 to 6.

Images