Strain modification method, system and equipment based on multi-objective optimization and medium

Abstract

The invention discloses a strain modification method, system and equipment based on multi-objective optimization and a medium, and the method comprises the steps: building a mathematical model of strain modification according to the task and demand of strain modification; according to the mathematical model of strain modification, a searching method of a strain modification target based on multi-objective optimization is provided; constructing a primary population according to a multi-target intelligent optimization task described by a mathematical model for strain modification; obtaining an Nth parent population; obtaining an Nth variation population according to crossover variation of the Nth parent population; calculating an updating performance index of the Nth variation population, and determining an (N + 1) th parent population according to the updating performance index; if the (N + 1) th parent population meets the preset condition, an optimal strain modification scheme is generated according to the (N + 1) th parent population, and multi-objective planning can be utilized to obtain strains capable of producing a large amount of target products and realizing steady-state growth, so that the yield of specific compounds is increased, and the production performance is enhanced.

Description

Technical Field

[0001] The present application relates to the technical field of optimized strain transformation, and in particular to a strain transformation method, system, equipment and medium based on multi-objective optimization. Background Art

[0002] With the continuous improvement of information technology, in modern biotechnology, the production of specific compounds can be increased by changing the natural metabolic pathways of microorganisms, or new pathways can be introduced to produce non-natural products. In the current strain modification technology, the existing genetic algorithm optimization only considers the influence of a single factor, which cannot achieve the best modification effect. In addition, different modification and cultivation of strains by existing technologies cannot make extensive modifications to the reaction network. The feasibility and efficiency of experimental optimization methods have technical problems such as long experimental cycles, unsatisfactory cultivation effects, and low efficiency. Summary of the invention

[0003] The following is a summary of the subject matter described in detail herein. This summary is not intended to limit the scope of the claims.

[0004] The main purpose of the embodiments of the present disclosure is to propose a strain transformation method, system, equipment and storage medium based on multi-objective optimization, which can optimize cell growth and growth-coupled synthesis, identify and screen high-quality optimization strategies, and improve production output and production performance.

[0005] A first aspect of an embodiment of the present application provides a strain transformation method based on multi-objective optimization, which is used in a central controller, and the method comprises:

[0006] According to the tasks and requirements of strain transformation, a mathematical model of the strain transformation is established;

[0007] According to the mathematical model of the strain transformation, a search method for strain transformation targets based on multi-objective optimization is proposed, so as to reduce the number of strain reactions as candidate variables in the multi-objective intelligent optimization task described by the mathematical model of the strain transformation according to the search method;

[0008] Constructing a primary population according to a multi-objective intelligent optimization task described by the mathematical model of strain transformation;

[0009] Obtaining the Nth parent population; if N is 1, the Nth parent population is selected from the initial population; if N is a positive integer greater than 1, the Nth parent population is generated based on the N-1th parent population;

[0010] According to the crossover mutation of the Nth parent population, an Nth mutation population is obtained;

[0011] Calculating the update performance index of the Nth mutant population, and determining the N+1th parent population according to the update performance index; if the N+1th parent population meets the preset conditions, generating the optimal strain transformation plan according to the N+1th parent population;

[0012] The process of selecting the Nth parent population from the primary population includes:

[0013] Randomly select two individuals of the initial population from the initial population; calculate the fitness of the individuals of the initial population respectively, and use the individual with the larger fitness as the individual of the Nth frontier parent population; wherein, if the fitness of the individuals of the two initial populations is the same, calculate the crowding distance of the individuals of the initial population respectively, and use the individual with the larger crowding distance as the individual of the Nth frontier parent population;

[0014] Calculating the fitness of all non-Nth frontier parent population individuals in the initial population respectively; calculating the selection probability of all non-Nth frontier parent population individuals in the initial population according to the fitness of non-Nth frontier parent population individuals in the initial population; using the roulette method, determining the individuals of the Nth subsequent parent population according to the selection probability of non-Nth frontier parent population individuals in the initial population;

[0015] The Nth frontier parent population and the Nth subsequent parent population are used as the Nth parent population.

[0016] The embodiment of the present application provides a strain modification method based on multi-objective optimization, by establishing a mathematical model for strain modification according to the tasks and requirements of the strain modification; proposing a method for searching for strain modification targets based on multi-objective optimization according to the mathematical model of the strain modification; constructing a primary population according to the multi-objective intelligent optimization tasks described by the mathematical model of the strain modification; obtaining the Nth parent population; obtaining the Nth variant population according to the crossover mutation of the Nth parent population; calculating the updated performance index of the Nth variant population, and determining the N+1th parent population according to the updated performance index; if the N+1th parent population meets the preset conditions, generating the optimal strain modification plan according to the N+1th parent population, and being able to use multi-objective planning to obtain strains that can both produce a large amount of target products and achieve steady-state growth, thereby increasing the yield of specific compounds and achieving enhanced production performance.

[0017] In some embodiments of the present application, the mathematical model of strain modification includes:

[0018] Assume that a strain has a metabolic network with m metabolites and n reactions, and the metabolic space of the metabolic network is defined as:

[0019] FS(v)={v∈R n |Sv=0,lb≤v≤ub};

[0020] Where FS(v) represents the metabolic space of v, S is the stoichiometric coefficient matrix of reaction m rows and n columns, v is the reaction rate vector on the metabolic network, and lb and ub are the range values ​​of v;

[0021] Construct two optimization objectives that satisfy the following conditions: maximize the yield of a given product P in the worst case max minw P , w P is the yield of product P after strain transformation and the coupling strength of maximizing yield and strain growth max min Cs:

[0022]

[0023]

[0024] u∈FS(u),w∈FS(w);

[0025]

[0026] in, is the production rate of the target metabolite, u and w represent the rate vectors of the wild type and perturbed steady state, respectively, F>1 is the user-defined multiple of the change between u and w, J is the set of modifiable reactions, is a binary 0-1 variable, representing the upregulation, downregulation and knockout operations of j∈J, respectively. K is the maximum number of preset modifications. The constraint condition It ensures that the candidate reaction can only perform one of the operations of upregulation, downregulation, and knockout. max or max(A(B=C)) refers to the maximum or minimum value of A when B=C. g and w p are the growth rate and production rate of the perturbed network under the considered design strategy, respectively. is the maximum growth rate.

[0027] In some embodiments of the present application, the method of reducing the number of strain responses as candidate variables in the multi-objective intelligent optimization task described by the mathematical model of strain transformation according to the search method includes:

[0028] According to the search method, eliminating strain reactions that do not meet preset conditions in the metabolic network;

[0029] Compressing the strain response to obtain candidate variables in a multi-objective intelligent optimization task described by a mathematical model of the strain transformation;

[0030] After proposing the method for searching targets for strain transformation based on multi-objective optimization, the method further includes:

[0031] The strain response is chromosomally encoded according to the metabolic network to obtain a plurality of strain chromosomes, wherein the strain chromosome contains at least one of the strain responses.

[0032] In some embodiments of the present application, the crossover mutation according to the Nth parent population to obtain the Nth mutant population includes:

[0033] Randomly select a chromosome position from the individuals of the Nth parent population as a crossover point;

[0034] Exchanging the chromosome segments of the individuals of the Nth parent population at the crossover point to obtain processed individuals of the Nth parent population;

[0035] Randomly selecting at least one chromosome position as a mutation point from the individuals of the Nth parent population after the treatment;

[0036] Mutating the chromosomes corresponding to the individuals of the processed Nth parent population at all the mutation points to obtain individuals of the Nth child population;

[0037] The Nth parent population is combined with the Nth child population to obtain the Nth variant population.

[0038] In some embodiments of the present application, the calculating the update performance index of the Nth mutant population and determining the N+1th parent population according to the update performance index includes:

[0039] Calculating target values ​​of all individuals in the Nth variant population according to the mathematical model of strain transformation;

[0040] Sort the individuals in the Nth variant population by non-dominance according to the target values ​​of the individuals in the Nth variant population, and determine the non-dominance level of each individual in the Nth variant population;

[0041] Calculating the crowding distance of each individual in the non-dominated level;

[0042] The individuals of the N+1th parent population are determined according to the non-dominated hierarchy of the individuals of the Nth variant population. When the number of individuals in the non-dominated hierarchy is greater than the size of the N+1th parent population, the individuals with the highest crowding distance in the non-dominated hierarchy are taken as the individuals of the N+1th parent population.

[0043] In some embodiments of the present application, the calculating the update performance index of the Nth mutant population and determining the N+1th parent population according to the update performance index includes:

[0044] Calculate target values ​​of all individuals in the Nth variant population according to the metabolic network;

[0045] Sort the individuals in the Nth variant population by non-dominance according to the target values ​​of the individuals in the Nth variant population, and determine the non-dominance level of each individual in the Nth variant population;

[0046] Calculating the crowding distance of each individual in the non-dominated level;

[0047] The individuals of the N+1th parent population are determined according to the non-dominated hierarchy of the individuals of the Nth variant population. When the number of individuals in the non-dominated hierarchy is greater than the size of the N+1th parent population, the individuals with the highest crowding distance in the non-dominated hierarchy are taken as the individuals of the N+1th parent population.

[0048] In some embodiments of the present application, the formula for calculating the crowding distance includes:

[0049]

[0050] CD first,n =CD end,n =1;

[0051]

[0052] Among them, y i-1,n represents the nth target value of the i-1th individual, y i+1,n represents the nth target value of the i+1th individual, y fifst,rn Represents the nth target value of the first individual in the current parent population, y end,n Indicates the nth target value of the last individual in the current parent population, CD i,n Represents the nth target value y of the ith individual i,n The crowding distance, CD i Represents the total value of the crowding distance in the i-th individual.

[0053] To achieve the above object, a second aspect of an embodiment of the present invention provides a strain transformation system based on multi-objective optimization, the system comprising:

[0054] Establishing a module for establishing a mathematical model of the strain transformation according to the tasks and requirements of the strain transformation;

[0055] An optimization module, for proposing a search method for strain transformation targets based on multi-objective optimization according to the mathematical model of strain transformation, so as to reduce the number of strain reactions as candidate variables in the multi-objective intelligent optimization task described by the mathematical model of strain transformation according to the search method;

[0056] A construction module, used to construct a primary population according to a multi-objective intelligent optimization task described by the mathematical model of the strain transformation;

[0057] An acquisition module, used for acquiring an Nth parent population; if N is 1, the Nth parent population is selected from the primary population; if N is a positive integer greater than 1, the Nth parent population is generated based on the N-1th parent population;

[0058] An iteration module, used for obtaining an Nth mutant population according to the crossover mutation of the Nth parent population;

[0059] An updating module, used to calculate an updating performance index of the Nth mutant population, and determine the N+1th parent population according to the updating performance index; if the N+1th parent population meets a preset condition, an optimal strain transformation plan is generated according to the N+1th parent population;

[0060] The process of selecting the Nth parent population from the primary population includes:

[0061] Randomly select two individuals of the initial population from the initial population; calculate the fitness of the individuals of the initial population respectively, and use the individual with the larger fitness as the individual of the Nth frontier parent population; wherein, if the fitness of the individuals of the two initial populations is the same, calculate the crowding distance of the individuals of the initial population respectively, and use the individual with the larger crowding distance as the individual of the Nth frontier parent population;

[0062] Calculating the fitness of all non-Nth frontier parent population individuals in the initial population respectively; calculating the selection probability of all non-Nth frontier parent population individuals in the initial population according to the fitness of non-Nth frontier parent population individuals in the initial population; using the roulette method, determining the individuals of the Nth subsequent parent population according to the selection probability of non-Nth frontier parent population individuals in the initial population;

[0063] The Nth frontier parent population and the Nth subsequent parent population are used as the Nth parent population.

[0064] To achieve the above-mentioned purpose, the third aspect of an embodiment of the present invention provides an electronic device, comprising: at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor so that the at least one control processor can execute the above-mentioned strain transformation method based on multi-objective optimization.

[0065] To achieve the above objectives, the fourth aspect of an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the above-mentioned strain transformation method based on multi-objective optimization.

[0066] It can be understood that the beneficial effects of the second to fourth aspects compared with the related art are the same as the beneficial effects of the first aspect compared with the related art. Please refer to the relevant description in the first aspect, and no further details will be given here. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0068] Figure 1 It is a schematic diagram of a process of a strain transformation method based on multi-objective optimization provided in an embodiment of the present application;

[0069] Figure 2 It is an iterative schematic diagram of a strain transformation method based on multi-objective optimization provided in an embodiment of the present application;

[0070] Figure 3 is the Pareto front of the final population generated by an implementation method provided in an embodiment of the present application;

[0071] Figure 4 It is a schematic diagram of the structure of a strain transformation system based on multi-objective optimization provided in an embodiment of the present application;

[0072] Figure 5 It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION

[0073] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be understood as limiting the present application.

[0074] In the description of this application, if there is a description of first, second, etc., it is only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0075] In the description of the present application, it should be understood that the descriptions involving orientation, such as the orientation or positional relationship indicated as up, down, etc., are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0076] In the description of this application, it should be noted that, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meaning of the above terms in this application based on the specific content of the technical solution.

[0077] First, some nouns involved in this application are analyzed:

[0078] In the field of strain cultivation design, multi-objective strain optimization refers to the use of multi-objective planning to achieve increased strain cultivation capacity and optimize strain cultivation effects.

[0079] Microbial cell factories (MCFs) are biological systems modified by modern biotechnology to efficiently produce target products, which mainly involves changing the natural metabolic pathways of microorganisms to increase the production of specific compounds, or introducing new pathways to produce non-natural products.

[0080] In actual industrial production, the feasibility of microbial cell factories requires extensive modification of the reaction network to enhance its production performance. However, in the current strain modification technology, the existing genetic algorithm optimization only considers the influence of a single factor and cannot achieve the best modification effect. In addition, different modification and cultivation methods are adopted for strains through existing technologies, and the reaction network cannot be extensively modified. The feasibility and efficiency of the experimental modification methods have technical problems such as long test cycles, unsatisfactory cultivation effects and low efficiency.

[0081] Based on this, the embodiments of the present application provide a strain modification method, system, electronic device and medium based on multi-objective optimization, which aims to transform cell growth and growth-coupled synthesis, identify and screen high-quality optimization strategies, and improve production yield and production performance.

[0082] The strain modification method, system, electronic device and medium based on multi-objective optimization provided in the embodiments of the present application are specifically illustrated by the following embodiments. First, the strain modification method based on multi-objective optimization in the embodiments of the present application is described.

[0083] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0084] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0085] The strain modification method based on multi-objective optimization provided in the embodiment of the present application relates to the technical field of strain modification. The strain modification method based on multi-objective optimization provided in the embodiment of the present application can be applied to a terminal, can also be applied to a server side, and can also be software running in a terminal or a server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc.; the server side can be configured as an independent physical server, or can be configured as a server cluster or a distributed system composed of multiple physical servers, and can also be configured as a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application for implementing strain modification based on multi-objective optimization, etc., but is not limited to the above forms.

[0086] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0087] It should be noted that in each specific implementation of the present application, when it comes to the need to perform relevant processing based on data related to user identity or characteristics such as user information, user behavior data, user historical data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of these data will comply with relevant laws, regulations, and standards. In addition, when the embodiment of the present application needs to obtain the user's sensitive personal information, the user's separate permission or consent will be obtained through a pop-up window or by jumping to a confirmation page. After clearly obtaining the user's separate permission or consent, the necessary user-related data for the normal operation of the embodiment of the present application will be obtained.

[0088] For this purpose, refer to Figure 1 The embodiment of the present application provides a strain transformation method based on multi-objective optimization. The method is applied to a central controller, which may be a server, an electronic device, or a mobile terminal, etc., and is not specifically limited here. The method includes the following steps S110 to S160:

[0089] Step S110: Establish a mathematical model for strain transformation according to the tasks and requirements of strain transformation.

[0090] In this step, first, according to the task and needs of strain transformation, laboratory techniques (such as mass spectrometry) are used to determine the metabolites produced by the strain under specific conditions. Existing bioinformatics databases (such as KEGG, MetaCyc, etc.) can also be used to query the possible metabolic pathways and the enzymes involved according to the strain name or genome information, obtain the strain metabolites and strain reactions, and then establish a mathematical model for strain transformation.

[0091] The mathematical model of strain transformation is explained in detail below:

[0092] Assume that a strain has a metabolic network with m metabolites and n reactions. The metabolic space of the metabolic network is defined as:

[0093] FS(v)={v∈R n |Sv=0,lb≤v≤ub};

[0094] Where FS(v) represents the metabolic space of v, S is the stoichiometric coefficient matrix of reaction m rows and n columns, v is the reaction rate vector on the metabolic network, and lb and ub are the range values ​​of v;

[0095] Construct two optimization objectives that satisfy the following conditions: maximize the yield of a given product P in the worst case max minw P , w P is the yield of product P after strain transformation and the coupling strength of maximizing yield and strain growth max min Cs:

[0096]

[0097] The coupling strength constraint is expressed as:

[0098]

[0099] The rate increase constraint is expressed as:

[0100]

[0101] The rate reduction constraint is expressed as:

[0102]

[0103] The reaction knockout constraint is expressed as:

[0104]

[0105] The feasible operation constraints are expressed as:

[0106]

[0107] The operation quantity constraint is expressed as:

[0108]

[0109] The flux space constraint is expressed as:

[0110] u∈FS(u),w∈FS(w);

[0111] The operation variable constraints are expressed as:

[0112]

[0113] in, is the production rate of the target metabolite, u and w represent the rate vectors of the wild type and perturbed steady state, respectively, F>1 is the user-defined multiple of the change between u and w, J is the set of modifiable reactions, is a binary 0-1 variable, representing the upregulation, downregulation and knockout operations of j∈J, respectively. K is the maximum number of preset modifications. The constraint condition It ensures that the candidate reaction can only perform one of the operations of upregulation, downregulation, and knockout. max or max(A(B=C)) refers to the maximum or minimum value of A when B=C. g and w p are the growth rate and production rate of the perturbed network under the considered design strategy, respectively. is the maximum growth rate.

[0114] Step S120: According to the mathematical model of strain transformation, a search method for strain transformation targets based on multi-objective optimization is proposed to reduce the number of strain responses as candidate variables in the multi-objective intelligent optimization task described by the mathematical model of strain transformation according to the search method.

[0115] In this step, the target of strain transformation refers to the reaction of the main target of the final strain transformation determined according to the mathematical model of strain transformation. According to the mathematical model of strain transformation, a search method for strain transformation targets based on multi-objective optimization is proposed, which specifically includes a comprehensive analysis of the metabolic network of the strain, identifying all potential controllable reactions, and then reducing the number of strain reactions as candidate variables in the multi-objective intelligent optimization task described by the mathematical model of strain transformation based on the identified controllable reactions and analysis results. This can not only effectively reduce the number of strain reactions as candidate variables in the multi-objective intelligent optimization task, but also improve the production efficiency of the target product while ensuring cell homeostasis.

[0116] In some embodiments, in step S120, reducing the number of strain responses as candidate variables in the multi-objective intelligent optimization task described by the mathematical model of strain transformation according to the search method includes the following steps:

[0117] Step S210: Eliminate strain reactions that do not meet preset conditions in the metabolic network according to the search method;

[0118] Step S220: compress the strain response to obtain candidate variables in the multi-objective intelligent optimization task described by the mathematical model of strain transformation.

[0119] In this embodiment, according to the controllable reactions identified by the search method for strain transformation targets based on multi-objective optimization and the analysis results, the number of strain reactions as candidate variables in the multi-objective intelligent optimization task described by the mathematical model of strain transformation is determined to be reduced. First, those strain reactions that are always inactive or have constant flux values ​​in the metabolic network, that is, strain reactions that do not affect cell homeostasis and product yield, are eliminated. Then, the linear reaction paths are compressed. Specifically, the situation where only one reaction product is the only substrate of another reaction can be merged to reduce the number of redundant reactions in the metabolic network, so that the model can be simplified to ensure that the stoichiometric matrix S of the original metabolic network is consistent with the simplified metabolic network, while reducing the number of elements in J and maintaining the consistency of the overall metabolic flow.

[0120] In some embodiments, after the search method for strain transformation targets based on multi-objective optimization is proposed in step S120, the following steps are also included:

[0121] The strain response is encoded by chromosomes according to the metabolic network to obtain a plurality of strain chromosomes, wherein the strain chromosome contains at least one strain response.

[0122] In one embodiment, the chromosome is first encoded, there are n regulatable reactions in the metabolic network, and the chromosome is defined as a vector containing n elements, each element corresponds to the operation of a reaction, and each reaction can be up-regulated, down-regulated, disabled, or not disturbed. Therefore, each element of the chromosome can take one of four operations: 0 (no disturbance), 1 (up-regulated), 2 (down-regulated), 3 (knockout), and by encoding the chromosome, the problem can be iteratively solved by the improved non-0-1 encoding genetic algorithm, and it is conducive to implementation on a computer device.

[0123] In this step, the strain response is chromosomally encoded according to the metabolic network, which is to convert each controllable strain response into a number or symbol for processing in the genetic algorithm, so that the problem can be iteratively solved by the improved non-0-1 encoding genetic algorithm, and is conducive to implementation on computer equipment.

[0124] Step S130: constructing a primary population according to the multi-objective intelligent optimization task described by the mathematical model of strain transformation.

[0125] Step S140, obtaining the Nth parent population; if N is 1, the Nth parent population is selected from the initial population; if N is a positive integer greater than 1, the Nth parent population is generated based on the N-1th parent population.

[0126] In this step, the first parent population is selected from the initial population. When N is a positive integer greater than 1, the Nth parent population is generated based on the N-1th parent population.

[0127] The following explains the process of selecting the Nth parent population from the initial population:

[0128] Randomly select two individuals from the initial population; calculate the fitness of the individuals in the initial population respectively, and take the individual with the larger fitness as the individual of the Nth frontier parent population; if the fitness of the individuals in the two initial populations is the same, calculate the crowding distance of the individuals in the initial population respectively, and take the individual with the larger crowding distance as the individual of the Nth frontier parent population;

[0129] Calculate the fitness of all non-Nth frontier parent population individuals in the initial population respectively; calculate the selection probability of all non-Nth frontier parent population individuals in the initial population according to the fitness of non-Nth frontier parent population individuals in the initial population; use the roulette method to determine the individuals of the Nth subsequent parent population according to the selection probability of non-Nth frontier parent population individuals in the initial population;

[0130] The Nth frontier parent population and the Nth subsequent parent population are taken as the Nth parent population.

[0131] In one embodiment, an improved non-dominated sorting genetic algorithm II (NSGA-II) is established to form an initial design population, and an initial evaluation is performed, and then an initial population is established by random generation. In addition, considering that it is difficult to modify a large number of reactions in a metabolic network in practical applications, the non-zero entries are set as At the same time, considering that knocking out too many reactions may lead to cell death, in order to increase the possibility of generating feasible solutions, the number of knockout reactions is limited to a maximum of 3, making the initially generated population reasonable and feasible.

[0132] Furthermore, in order to select individuals for crossover mutation, a combination of roulette wheel selection and tournament selection is used. The Nth frontier parent population represents the parent population selected in the first 20% of iterations, and the 2-tournament selection method is applied to select according to the ranking of chromosomes in the non-dominated sorting; if a tie occurs, the crowding distance of the chromosomes is compared. The Nth subsequent parent population represents the parent population selected in the subsequent 80% of iterations, and roulette wheel selection is used, based on the fitness of the chromosome (defined as the crowding distance of the chromosome), and the fitness of the frontier individuals is set to 1.

[0133] By combining tournament and roulette in the selection of crossover individuals, high-quality chromosomes have a higher crossover probability, thereby guiding the population iteration towards a better solution. At the same time, it is set that there is at least one non-zero item at the crossover site to ensure the effectiveness of the crossover.

[0134] The following explanation of the formula for calculating crowding distance includes:

[0135]

[0136] CD first,n =CD end,n =1;

[0137]

[0138] Among them, y i-1,n represents the nth target value of the i-1th individual, y i+1,n represents the nth target value of the i+1th individual, y first,n Represents the nth target value of the first individual in the current parent population, y end,n Indicates the nth target value of the last individual in the current parent population, CD i,n Represents the nth target value y of the ith individual i,n The crowding distance, CDi Represents the total value of the crowding distance in the i-th individual.

[0139] In one embodiment, each individual in the initial population is individually evaluated based on fitness value (i.e., cell growth rate and yield of target reaction), and combined with crowding distance calculation and non-dominated sorting, the results are used for pair selection to perform crossover operation.

[0140] Step S150: performing crossover mutation according to the Nth parent population to obtain the Nth mutant population.

[0141] In this step, firstly, the Nth parent population is cross-mutated to obtain the Nth child population, and then the Nth parent population is combined with the Nth child population to obtain the Nth mutant population.

[0142] In some embodiments, in step S150, the Nth mutant population is obtained according to the crossover mutation of the Nth parent population, including the following steps:

[0143] Step S310, randomly selecting a chromosome position from the individuals in the Nth parent population as a crossover point;

[0144] Step S320, exchanging the chromosome segments of the individuals of the Nth parent population at the crossover point to obtain the processed individuals of the Nth parent population;

[0145] Step S330: randomly select at least one chromosome position from the processed individuals of the Nth parent population as a mutation point, and mutate the chromosomes corresponding to the processed individuals of the Nth parent population at all mutation points to obtain individuals of the Nth child population;

[0146] Step S340: Combine the Nth parent population with the Nth child population to obtain the Nth mutant population.

[0147] In this embodiment, the crossover operation adopts a simple single-point crossover method, randomly selecting a chromosome position from the individuals of the Nth parent population as a crossover point, and exchanging the chromosome fragments of the individuals of the Nth parent population at the crossover point to obtain the processed individuals of the Nth parent population. During the crossover, it is ensured that at least one non-zero point is involved in the crossover to ensure the effectiveness of the crossover operation.

[0148] Furthermore, the mutation operation uses multi-point mutation to generate mutant chromosomes, randomly selects at least one chromosome position from the individuals of the processed Nth parent population as a mutation point, mutates the chromosomes corresponding to the individuals of the processed Nth parent population at all mutation points, and obtains individuals of the Nth child population; then, the Nth parent population is combined with the Nth child population to obtain the Nth mutant population.

[0149] In one embodiment, the process of mutating the chromosomes corresponding to the individuals of the processed Nth parent population in step S330 includes the following steps:

[0150] Step S410, randomly generating a first variation value;

[0151] Step S420: When the first mutation value is less than or equal to the preset mutation probability, the chromosome corresponding to the individual of the processed Nth parent population is replaced with the second mutation value to obtain the individual of the Nth parent population after mutation; when the first mutation value is greater than the preset mutation probability, the individual of the Nth parent population after mutation is used as the individual of the Nth parent population after mutation, wherein the second mutation value is a randomly generated value within a preset range;

[0152] Step S430, determining the non-zero elements of individuals in the Nth parent population after mutation according to the individuals in the Nth parent population after mutation;

[0153] Step S440: When the number of non-zero elements is greater than a preset number, randomly replace at least one non-zero element in the individuals of the Nth parent population after mutation with a zero element until the number of non-zero elements is less than or equal to the preset number, so as to obtain individuals of the Nth child population;

[0154] Step S450: When the number of non-zero elements is less than or equal to a preset number, the individuals of the Nth parent population after mutation are used as individuals of the Nth child population.

[0155] In this embodiment, the first mutation value is preferably a random number in the range of 0 to 1. When the first mutation value is less than or equal to the preset mutation probability, the chromosome corresponding to the individual of the processed Nth parent population is replaced with the second mutation value to obtain the individual of the Nth parent population after mutation; when the first mutation value is greater than the preset mutation probability, the individual of the Nth parent population after mutation is used as the individual of the Nth parent population after mutation, wherein the second mutation value is preferably a random integer between 0 and 3. For the new individual obtained after mutation, the number of non-zero elements on its chromosome is counted. If the number of non-zero elements exceeds the preset maximum number limit, at least one non-zero element is randomly selected and set to 0 until the number of non-zero elements meets the requirement. When the number of non-zero elements is less than or equal to the preset number, the individual of the Nth parent population after mutation is used as the individual of the Nth child population to ensure that the mutation operation does not destroy the basic structure of the chromosome.

[0156] In one embodiment, a randomly generated number r (in the range of [0,1]) is compared with a predefined mutation probability Pm. If r≤Pm, the corresponding position will mutate, and the original value will be replaced by a random integer between 0 and 3. The feasibility of the mutated solution is checked. In particular, if the number of non-zero elements in the mutated solution exceeds K, some non-zero elements are randomly selected and reset to zero until the total number of non-zero elements does not exceed K. After the mutation, the original population and the mutated individuals are combined, and further operations such as non-dominated sorting and environmental selection are performed.

[0157] Step S160, calculating the update performance index of the Nth mutant population, and determining the N+1th parent population according to the update performance index; if the N+1th parent population meets the preset conditions, generating the optimal strain transformation plan according to the N+1th parent population.

[0158] In this step, the target values ​​of all individuals in the Nth variant population are calculated according to the metabolic network, the individuals in the Nth variant population are sorted by non-dominance according to the target values ​​of the individuals in the Nth variant population, the non-dominance level of the individuals in each Nth variant population is determined, the crowding distance of the individuals in each non-dominance level is calculated, and the individuals of the N+1th parent population are determined according to the non-dominance level of the individuals in the Nth variant population. When the number of individuals in the non-dominance level is greater than the size of the N+1th parent population, the individual with the highest crowding distance of the individuals in the non-dominance level is taken as the individual of the N+1th parent population.

[0159] In some embodiments, calculating the update performance index of the Nth mutant population in step S160 and determining the N+1th parent population according to the update performance index includes the following steps:

[0160] Step S510, calculating the target values ​​of all individuals in the Nth variant population according to the metabolic network;

[0161] Step S520, sorting the individuals in the Nth variant population according to the target values ​​of the individuals in the Nth variant population, and determining the non-dominated level of the individuals in each Nth variant population;

[0162] Step S530, calculating the crowding distance of individuals in each non-dominated level;

[0163] Step S540: determine the individuals of the N+1th parent population according to the non-dominated hierarchy of the individuals of the Nth mutant population. When the number of individuals in the non-dominated hierarchy is greater than the size of the N+1th parent population, the individual with the highest crowding distance in the non-dominated hierarchy is used as the individual of the N+1th parent population.

[0164] In this embodiment, non-dominated degree sorting and crowded degree sorting are constructed and executed, the individuals in the population that have undergone crossover mutation are evaluated, and excellent individuals are screened to form new offspring. The target values ​​of individuals in the mutant population are calculated for quantitative evaluation and consideration, providing data support for further screening to form new parent generations.

[0165] For the nth target value y of the ith individual in any parent population i,n , use the following formula to calculate the individual crowding distance:

[0166]

[0167] CD first,n =CD end,n =1;

[0168]

[0169] Among them, y i-1,n represents the nth target value of the i-1th individual, y i+1,n represents the nth target value of the i+1th individual, y first,n Represents the nth target value of the first individual in the current parent population, y end,n Indicates the nth target value of the last individual in the current parent population, CD i,n Represents the nth target value y of the ith individual i,n The crowding distance, CD i Represents the total value of the crowding distance in the i-th individual. The crowding value is used to distinguish individuals on the same frontier. Individuals with high values ​​are relatively more dispersed, and more diverse individuals can be given priority during selection.

[0170] Among them, the dominance relationship of individual x1 to another individual x2 means that the target value of x1 is greater than the target value of x2. Therefore, after determining the dominance relationship, the individual that is not dominated by any other individual is classified as the first frontier individual. The individual that is only dominated by the first frontier individual is classified as the second frontier individual, and so on, completing the frontier allocation of all individuals, and then the individuals in each frontier are arranged in descending order according to the target value.

[0171] Finally, individuals are selected based on the non-dominated frontier level. When the population size exceeds the set upper limit, individuals with higher crowding are selected to form a new parent generation and enter the next cycle until the iteration requirements are met. The iteration requirements are not limited here.

[0172] In one embodiment of this example, the method proposed in the present invention is illustrated in conjunction with the case of overproduction of succinic acid in Escherichia coli. We aim to apply our method to this task in order to identify a reasonable design strategy to achieve high succinic acid production while ensuring the survival and growth of Escherichia coli.

[0173] The present invention proposes a design method for a coupled cell strain based on multi-objective transformation, comprising the following steps:

[0174] First, a metabolic model of E. coli is established. In this case, the latest E. coli genome-scale metabolic network model iML1515 is used. At the same time, a multi-objective mathematical optimization model is established as follows:

[0175]

[0176] u∈FS(u),w∈FS(w);

[0177]

[0178] in, is the production rate of succinate, u and w represent the flux vectors of the wild type and perturbed steady state, respectively, F>1 is the change factor between u and w defined by the user, which is set to 2 in this implementation, and J is the set of modifiable reactions. is a binary 0-1 variable, representing the upregulation, downregulation and knockout operations of j∈J respectively. K represents the maximum value of the number of modifications set manually, which is set to 10 in this implementation.

[0179] Furthermore, a simplified model was established to remove reactions that were always inactive or had constant flux values ​​in the metabolic network, and to compress linear reactions. The iML1515 model contains 1515 genes connected to 2719 reactions, involving 1192 metabolites. After model compression and candidate reaction screening, we obtained 352 aggregate reactions as interference candidates. These 352 aggregate reactions constitute set J.

[0180] Further, the NSGA-II algorithm is run. The flowchart of the NSGA-II algorithm is as follows: Figure 2 As shown, in this embodiment, the population size is set to 100, and the transformation is performed for a maximum of 100 generations. The crossover rate is set to 0.9, and the mutation rate is 1 / L, where L is the total number of binary decision variables. The population is initialized and meets And remove the initial conditions that do not exceed three, and then perform the initial non-domination and crowding sorting to facilitate the initial crossover operation.

[0181] Furthermore, a crossover mutation operation is performed, and the individuals that undergo crossover are calculated based on the crossover probability. Individual selection is performed by combining tournament and roulette. In the first 20% of iterations, the 2-tournament selection method is applied to select individuals based on the ranking of chromosomes in the non-dominated sorting; if a tie occurs, the crowding distance of the chromosomes is compared. In the subsequent 80% of iterations, roulette selection is used, and the fitness of the border individuals is set to 1 based on the fitness of the chromosomes (defined as the crowding distance of the chromosomes). Single-point crossover occurs, and the validity of the crossover is guaranteed, and 0-0 crossover is prohibited.

[0182] Furthermore, the mutation probability Pm is set to 1 / L, and the randomly generated number r (in the range [0,1]) is compared with the predefined mutation probability Pm. If r≤Pm, the corresponding position will mutate, and the original value will be replaced by a random integer between 0 and 3. The feasibility of the mutated solution is checked. In particular, if the number of non-zero elements in the mutated solution exceeds K, some non-zero elements are randomly selected and reset to zero until the total number of non-zero elements does not exceed K. After mutation, the original population and the mutated individuals are combined, and further operations such as non-dominated sorting and environmental selection are performed.

[0183] Furthermore, after the non-dominated degree is sorted and the dominance relationship is determined, the individuals that are not dominated by any other individuals are classified as first-front individuals, and the individuals that are only dominated by the first-front individuals are classified as second-front individuals, and so on, completing the frontier allocation of all individuals, and then sorting the individuals on each front in descending order according to the target value.

[0184] Finally, the population evolution is selected according to the non-dominated frontier level. According to the iteration condition, when the population size exceeds the set upper limit, individuals with higher crowding degree are selected to form a new parent generation and enter the next cycle until 100 iterations are completed. The Pareto frontier of the final population generated is as follows: Figure 3 As shown, the modification status of the excellent individuals finally selected is shown in Table 1.

[0185] Table 1

[0186] result Increase Downward adjustment Knockout Modification times 1 RBF \ R5P 2 2 DHORD5 ALLTN,DHORTS ACHBS,FLVR,PETNT181pp,R5PP 7

[0187] In this embodiment, the strain is first modeled, a metabolic network model is established, and an improved non-dominated sorting genetic algorithm is introduced. Through mathematical modeling, a multi-objective optimization problem model of this type of design is established. The initialization process emphasizes rationality and feasibility, which not only ensures the survival of individuals, but also conforms to reality, and can realize the identification and screening of high-quality optimization strategies. Different from the traditional crossover individual selection method, the crossover individual selection method combining the design tournament and roulette is accelerated to accelerate the convergence speed of the population. According to the results of non-dominated degree sorting and crowding degree sorting, individuals are selected to become new parents, so that the population is continuously transformed in the direction of the optimal solution, while maintaining a certain population diversity, and product production can be coupled with growth, and corresponding individual modification strategies are provided to maintain the steady state of cells while increasing production. It has a wide range of applications in strain design, reflecting its prospects in designing high-yield strains in the field of biomanufacturing.

[0188] like Figure 4 As shown, some embodiments of the present application provide a strain transformation system based on multi-objective optimization, the system includes an establishment module 410, an optimization module 420, a construction module 430, an acquisition module 440, an iteration module 450, and an update module 460, specifically:

[0189] Establishing module 410, for establishing a mathematical model of strain transformation according to the tasks and requirements of strain transformation;

[0190] An optimization module 420 is used to propose a search method for strain transformation targets based on multi-objective optimization according to the mathematical model of strain transformation, so as to reduce the number of strain responses as candidate variables in the multi-objective intelligent optimization task described by the mathematical model of strain transformation according to the search method;

[0191] A construction module 430 is used to construct a primary population according to a multi-objective intelligent optimization task described by a mathematical model of strain transformation;

[0192] The acquisition module 440 is used to acquire the Nth parent population; if N is 1, the Nth parent population is selected from the initial population; if N is a positive integer greater than 1, the Nth parent population is generated based on the N-1th parent population;

[0193] Iteration module 450, used for performing crossover mutation according to the Nth parent population to obtain the Nth mutation population;

[0194] The updating module 460 is used to calculate the updating performance index of the Nth mutant population, and determine the N+1th parent population according to the updating performance index; if the N+1th parent population meets the preset conditions, the optimal transformation plan of the strain is generated according to the N+1th parent population;

[0195] The process of selecting the Nth parent population from the initial population includes:

[0196] Randomly select two individuals from the initial population; calculate the fitness of the individuals in the initial population respectively, and take the individual with the larger fitness as the individual of the Nth frontier parent population; if the fitness of the individuals in the two initial populations is the same, calculate the crowding distance of the individuals in the initial population respectively, and take the individual with the larger crowding distance as the individual of the Nth frontier parent population;

[0197] Calculate the fitness of all non-Nth frontier parent population individuals in the initial population respectively; calculate the selection probability of all non-Nth frontier parent population individuals in the initial population according to the fitness of non-Nth frontier parent population individuals in the initial population; use the roulette method to determine the individuals of the Nth subsequent parent population according to the selection probability of non-Nth frontier parent population individuals in the initial population;

[0198] The Nth frontier parent population and the Nth subsequent parent population are taken as the Nth parent population.

[0199] In some implementations, the establishing module 410 may include:

[0200] Assume that a strain has a metabolic network with m metabolites and n reactions. The metabolic space of the metabolic network is defined as:

[0201] FS(v)={v∈R n |Sv=0,lb≤v≤ub};

[0202] Where FS(v) represents the metabolic space of v, S is the stoichiometric coefficient matrix of reaction m rows and n columns, v is the reaction rate vector on the metabolic network, and lb and ub are the range values ​​of v;

[0203] Construct two optimization objectives that satisfy the following conditions: maximize the yield of a given product P in the worst case max minw P , w P is the yield of product P after strain transformation and the coupling strength of maximizing yield and strain growth max min Cs:

[0204]

[0205] u∈FS(u),w∈FS(w);

[0206]

[0207] in, is the production rate of the target metabolite, u and w represent the rate vectors of the wild-type and perturbed steady-state reactions, respectively, F>1 is the user-defined multiple of the change between u and w, and J is the set of modifiable reactions. is a binary 0-1 variable, representing the upregulation, downregulation and knockout operations of j∈J, respectively. K is the maximum number of preset modifications. The constraint condition It ensures that the candidate reaction can only perform one of the operations of upregulation, downregulation, and knockout. max or max(A(B=C)) refers to the maximum or minimum value of A when B=C. g and w p are the growth rate and production rate of the perturbed network under the considered design strategy, respectively. is the maximum growth rate.

[0208] In some embodiments, the optimization module 420 may include: eliminating strain responses that do not meet preset conditions in the metabolic network according to a search method.

[0209] In some embodiments, the optimization module 420 may include: compressing the strain responses to obtain candidate variables in a multi-objective intelligent optimization task described by a mathematical model of strain modification.

[0210] In some embodiments, the optimization module 420 may include: performing chromosome encoding on the strain response according to the metabolic network to obtain a plurality of strain chromosomes, wherein the strain chromosome includes at least one strain response.

[0211] In some embodiments, the iteration module 450 may include: randomly selecting a chromosome position from individuals in the Nth parent population as a crossover point.

[0212] In some implementations, the iteration module 450 may include: exchanging chromosome segments of individuals of the Nth parent population at a crossover point to obtain processed individuals of the Nth parent population.

[0213] In some implementations, the iteration module 450 may include: randomly selecting at least one chromosome position as a variation point from the processed individuals of the Nth parent population.

[0214] In some implementations, the iteration module 450 may include: mutating the chromosomes corresponding to the processed individuals of the Nth parent population at all mutation points to obtain individuals of the Nth child population.

[0215] In some implementations, the iteration module 450 may include: combining the Nth parent population with the Nth child population to obtain the Nth mutant population.

[0216] In some implementations, the iteration module 450 may include: randomly generating a first variation value.

[0217] In some embodiments, the iteration module 450 may include: when the first variation value is less than or equal to the preset variation probability, replacing the chromosome corresponding to the individual of the processed Nth parent population with the second variation value to obtain the individual of the mutated Nth parent population; when the first variation value is greater than the preset variation probability, using the individual of the processed Nth parent population as the individual of the mutated Nth parent population, wherein the second variation value is a randomly generated value within a preset range.

[0218] In some implementations, the iteration module 450 may include: determining non-zero elements of individuals in the Nth parent population after mutation according to the individuals in the Nth parent population after mutation.

[0219] In some embodiments, the iteration module 450 may include: when the number of non-zero elements is greater than a preset number, randomly replacing at least one non-zero element in the individuals of the Nth parent population after mutation with a zero element, until the number of non-zero elements is less than or equal to the preset number, to obtain individuals of the Nth child population.

[0220] In some embodiments, the iteration module 450 may include: when the number of non-zero elements is less than or equal to a preset number, using the individuals of the Nth parent population after mutation as the individuals of the Nth child population.

[0221] In some embodiments, the updating module 460 may include: calculating target values ​​of all individuals in the Nth variant population according to the metabolic network.

[0222] In some implementations, the updating module 460 may include: sorting the non-dominated degrees of the individuals in the Nth variant population according to the target values ​​of the individuals in the Nth variant population, and determining the non-dominated level of the individuals in each Nth variant population.

[0223] In some implementations, the updating module 460 may include calculating a crowding distance of individuals in each non-dominated level.

[0224] In some embodiments, the update module 460 may include: determining the individuals of the N+1th parent population based on the non-dominated hierarchy of the individuals of the Nth mutant population, and when the number of individuals in the non-dominated hierarchy is greater than the size of the N+1th parent population, taking the individual with the highest crowding distance among the individuals in the non-dominated hierarchy as the individual of the N+1th parent population.

[0225] In some implementations, the update module 460 may include:

[0226]

[0227] CD first,n =CD end,n =1;

[0228]

[0229] Among them, y i-1,n represents the nth target value of the i-1th individual, y i+1,n represents the nth target value of the i+1th individual, y first,n Represents the nth target value of the first individual in the current parent population, y end,n Indicates the nth target value of the last individual in the current parent population, CD i,n Represents the nth target value y of the ith individual i,n The crowding distance, CD i Represents the total value of the crowding distance in the i-th individual.

[0230] It should be noted that the strain transformation system based on multi-objective optimization provided in this embodiment and the above-mentioned strain transformation method based on multi-objective optimization are based on the same inventive concept. Therefore, the relevant content of the above-mentioned strain transformation method based on multi-objective optimization is also applicable to the content of the strain transformation system based on multi-objective optimization. Therefore, it will not be repeated here.

[0231] In order to solve the technical problems of the existing technology, such as long time consumption, poor transformation effect and low efficiency, the system establishes a mathematical model for strain transformation according to the tasks and requirements of strain transformation; according to the mathematical model of strain transformation, a search method for strain transformation targets based on multi-objective optimization is proposed; the initial population is constructed according to the multi-objective intelligent optimization tasks described by the mathematical model of strain transformation; the Nth parent population is obtained; the Nth mutant population is obtained according to the cross-mutation of the Nth parent population; the updated performance index of the Nth mutant population is calculated, and the N+1th parent population is determined according to the updated performance index; if the N+1th parent population meets the preset conditions, the optimal strain transformation plan is generated according to the N+1th parent population, and multi-objective planning can be used to obtain strains that can both produce a large amount of target products and achieve steady-state growth, thereby increasing the yield of specific compounds and achieving enhanced production performance.

[0232] An embodiment of the present application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the above-mentioned strain transformation method based on multi-objective optimization is implemented.

[0233] like Figure 5 , Figure 5 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application, wherein the electronic device includes:

[0234] At least one battery;

[0235] at least one memory;

[0236] at least one processor;

[0237] at least one program;

[0238] The program is stored in the memory, and the processor executes at least one program to implement the above-mentioned strain transformation method based on multi-objective optimization implemented in the present disclosure.

[0239] The electronic device may be any intelligent terminal including a mobile phone, a tablet computer, a personal digital assistant (PDA), a vehicle-mounted computer, etc.

[0240] The electronic device according to the embodiment of the present application is described in detail below.

[0241] The processor 1600 may be implemented by a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present disclosure;

[0242] The memory 1700 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1700 can store an operating system and other application programs. When the technical solution provided in the embodiments of this specification is implemented by software or firmware, the relevant program code is stored in the memory 1700, and the processor 1600 is called to execute a strain transformation method based on multi-objective optimization in the embodiment of the present disclosure.

[0243] Input / output interface 1800, used to implement information input and output;

[0244] The communication interface 1900 is used to realize the communication interaction between the device and other devices. The communication can be realized through a wired manner (such as USB, network cable, etc.) or a wireless manner (such as mobile network, WIFI, Bluetooth, etc.);

[0245] Bus 2000 , which transmits information between various components of the device (e.g., processor 1600 , memory 1700 , input / output interface 1800 , and communication interface 1900 );

[0246] The processor 1600 , the memory 1700 , the input / output interface 1800 , and the communication interface 1900 are connected to each other in communication within the device via the bus 2000 .

[0247] The embodiments of the present disclosure also provide a storage medium, which is a computer-readable storage medium. The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the above-mentioned strain transformation method based on multi-objective optimization.

[0248] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0249] The embodiments described in the embodiments of the present disclosure are intended to more clearly illustrate the technical solutions of the embodiments of the present disclosure and do not constitute a limitation on the technical solutions provided by the embodiments of the present disclosure. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of the present disclosure are also applicable to similar technical problems.

[0250] Those skilled in the art will appreciate that the technical solutions shown in the figures do not limit the embodiments of the present disclosure and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.

[0251] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0252] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0253] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0254] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0255] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0256] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0257] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0258] If the integrated unit 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 technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions to enable an electronic device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of each embodiment method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store programs.

[0259] The above is a specific description of the preferred implementation of the embodiments of the present application, but the embodiments of the present application are not limited to the above-mentioned implementation methods. Technical personnel familiar with the field can also make various equivalent modifications or substitutions without violating the spirit of the embodiments of the present application. These equivalent modifications or substitutions are all included in the scope defined by the claims of the embodiments of the present application.

[0260] The embodiments of the present application are described in detail above in conjunction with the accompanying drawings, but the present application is not limited to the above embodiments. Various changes can be made within the knowledge scope of ordinary technicians in the relevant technical field without departing from the purpose of the present application.

Claims

1. A strain transformation method based on multi-objective optimization, characterized in that: The method comprises: According to the tasks and requirements of strain transformation, a mathematical model of the strain transformation is established; According to the mathematical model of the strain transformation, a search method for strain transformation targets based on multi-objective optimization is proposed, so as to reduce the number of strain reactions as candidate variables in the multi-objective intelligent optimization task described by the mathematical model of the strain transformation according to the search method; Constructing a primary population according to a multi-objective intelligent optimization task described by the mathematical model of strain transformation; Obtaining the Nth parent population; if N is 1, the Nth parent population is selected from the initial population; if N is a positive integer greater than 1, the Nth parent population is generated based on the N-1th parent population; According to the crossover mutation of the Nth parent population, an Nth mutation population is obtained; Calculating the update performance index of the Nth mutant population, and determining the N+1th parent population according to the update performance index; if the N+1th parent population meets the preset conditions, generating the optimal strain transformation plan according to the N+1th parent population; The process of selecting the Nth parent population from the primary population includes: Randomly select two individuals of the initial population from the initial population; calculate the fitness of the individuals of the initial population respectively, and use the individual with the larger fitness as the individual of the Nth frontier parent population; wherein, if the fitness of the individuals of the two initial populations is the same, calculate the crowding distance of the individuals of the initial population respectively, and use the individual with the larger crowding distance as the individual of the Nth frontier parent population; Calculating the fitness of all non-Nth frontier parent population individuals in the initial population respectively; calculating the selection probability of all non-Nth frontier parent population individuals in the initial population according to the fitness of non-Nth frontier parent population individuals in the initial population; using the roulette method, determining the individuals of the Nth subsequent parent population according to the selection probability of non-Nth frontier parent population individuals in the initial population; The Nth frontier parent population and the Nth subsequent parent population are used as the Nth parent population.

2. The strain transformation method based on multi-objective optimization according to claim 1, characterized in that: The mathematical model of strain transformation includes: Assume that a strain has a metabolic network with m metabolites and n reactions, and the metabolic space of the metabolic network is defined as: FS(v)={v∈R n |Sv=0,lb≤v≤ub}; Where FS(v) represents the metabolic space of v, S is the stoichiometric coefficient matrix of reaction m rows and n columns, v is the reaction rate vector on the metabolic network, and lb and ub are the range values ​​of v; Construct two optimization objectives that satisfy the following conditions: maximize the yield of a given product P in the worst case max min w P , w P is the yield of product P after strain transformation and the coupling strength of maximizing yield and strain growth max min Cs: u∈FS(u),w∈FS(w); in, is the production rate of the target metabolite, u and w represent the rate vectors of the wild-type and perturbed steady-state reactions, respectively, F>1 is the user-defined multiple of the change between u and w, and J is the set of modifiable reactions. is a binary 0-1 variable, representing the upregulation, downregulation and knockout operations of j∈J, respectively. K is the maximum number of preset modifications. The constraint condition It ensures that the candidate reaction can only perform one of the operations of upregulation, downregulation, and knockout. max or max(A(B=C)) refers to the maximum or minimum value of A when B=C. g and w p are the growth rate and production rate of the perturbed network under the considered design strategy, respectively. is the maximum growth rate.

3. The strain transformation method based on multi-objective optimization according to claim 2, characterized in that: The method of reducing the number of strain responses as candidate variables in the multi-objective intelligent optimization task described by the mathematical model of strain transformation according to the search method includes: According to the search method, eliminating strain reactions that do not meet preset conditions in the metabolic network; Compressing the strain response to obtain candidate variables in a multi-objective intelligent optimization task described by a mathematical model of the strain transformation; After proposing the method for searching targets for strain transformation based on multi-objective optimization, the method further includes: The strain response is chromosomally encoded according to the metabolic network to obtain a plurality of strain chromosomes, wherein the strain chromosome contains at least one of the strain responses.

4. The strain transformation method based on multi-objective optimization according to claim 1, characterized in that: The step of obtaining the Nth mutant population according to the crossover mutation of the Nth parent population includes: Randomly select a chromosome position from the individuals of the Nth parent population as a crossover point; Exchanging the chromosome segments of the individuals of the Nth parent population at the crossover point to obtain processed individuals of the Nth parent population; Randomly selecting at least one chromosome position as a mutation point from the individuals of the Nth parent population after the treatment; Mutating the chromosomes corresponding to the individuals of the processed Nth parent population at all the mutation points to obtain individuals of the Nth child population; The Nth parent population is combined with the Nth child population to obtain the Nth variant population.

5. The strain transformation method based on multi-objective optimization according to claim 4, characterized in that: The process of mutating the chromosomes corresponding to the individuals of the processed Nth parent population includes: Randomly generate a first mutation value; When the first mutation value is less than or equal to the preset mutation probability, the chromosome corresponding to the individual of the Nth parent population after the processing is replaced with the second mutation value to obtain the individual of the Nth parent population after the mutation; when the first mutation value is greater than the preset mutation probability, the individual of the Nth parent population after the processing is used as the individual of the Nth parent population after the mutation, wherein the second mutation value is a randomly generated value within a preset range; Determine the non-zero elements of individuals in the Nth parent population after the mutation according to the individuals in the Nth parent population after the mutation; When the number of the non-zero elements is greater than the preset number, at least one non-zero element in the individuals of the Nth parent population after the mutation is randomly replaced with a zero element until the number of the non-zero elements is less than or equal to the preset number, so as to obtain individuals of the Nth child population; When the number of the non-zero elements is less than or equal to a preset number, the individuals of the Nth parent population after the mutation are used as the individuals of the Nth child population.

6. The strain transformation method based on multi-objective optimization according to claim 1, characterized in that: The calculating the update performance index of the Nth mutant population and determining the N+1th parent population according to the update performance index comprises: Calculating target values ​​of all individuals in the Nth variant population according to the mathematical model of strain transformation; Sort the individuals in the Nth variant population by non-dominance according to the target values ​​of the individuals in the Nth variant population, and determine the non-dominance level of each individual in the Nth variant population; Calculating the crowding distance of each individual in the non-dominated level; The individuals of the N+1th parent population are determined according to the non-dominated hierarchy of the individuals of the Nth variant population. When the number of individuals in the non-dominated hierarchy is greater than the size of the N+1th parent population, the individuals with the highest crowding distance in the non-dominated hierarchy are taken as the individuals of the N+1th parent population.

7. The strain transformation method based on multi-objective optimization according to claim 6, characterized in that: The formula for calculating the crowding distance includes: CD first,n =CD end,n =1; Among them, y i-1，n represents the nth target value of the i-1th individual, y i+1，n represents the nth target value of the i+1th individual, y first,n Represents the nth target value of the first individual in the current parent population, y end,n Indicates the nth target value of the last individual in the current parent population, CD i,n Represents the nth target value y of the ith individual i，n The crowding distance, CD i Represents the total value of the crowding distance in the i-th individual.

8. A strain transformation system based on multi-objective optimization, characterized in that: The system comprises: Establishing a module for establishing a mathematical model of the strain transformation according to the tasks and requirements of the strain transformation; An optimization module, for proposing a search method for strain transformation targets based on multi-objective optimization according to the mathematical model of strain transformation, so as to reduce the number of strain reactions as candidate variables in the multi-objective intelligent optimization task described by the mathematical model of strain transformation according to the search method; A construction module, used to construct a primary population according to a multi-objective intelligent optimization task described by the mathematical model of the strain transformation; An acquisition module, used for acquiring an Nth parent population; if N is 1, the Nth parent population is selected from the primary population; if N is a positive integer greater than 1, the Nth parent population is generated based on the N-1th parent population; An iteration module, used for obtaining an Nth mutant population according to the crossover mutation of the Nth parent population; An updating module, used to calculate an updating performance index of the Nth mutant population, and determine the N+1th parent population according to the updating performance index; if the N+1th parent population meets a preset condition, an optimal strain transformation plan is generated according to the N+1th parent population; The process of selecting the Nth parent population from the primary population includes: Randomly select two individuals of the initial population from the initial population; calculate the fitness of the individuals of the initial population respectively, and use the individual with the larger fitness as the individual of the Nth frontier parent population; wherein, if the fitness of the individuals of the two initial populations is the same, calculate the crowding distance of the individuals of the initial population respectively, and use the individual with the larger crowding distance as the individual of the Nth frontier parent population; Calculating the fitness of all non-Nth frontier parent population individuals in the initial population respectively; calculating the selection probability of all non-Nth frontier parent population individuals in the initial population according to the fitness of non-Nth frontier parent population individuals in the initial population; using the roulette method, determining the individuals of the Nth subsequent parent population according to the selection probability of non-Nth frontier parent population individuals in the initial population; The Nth frontier parent population and the Nth subsequent parent population are used as the Nth parent population.

9. An electronic device, characterized in that: It includes at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor so that the at least one control processor can execute a strain transformation method based on multi-objective optimization as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the strain transformation method based on multi-objective optimization described in any one of claims 1 to 7.

Images