Foraging algorithm for coupling dominant functional bacteria of targeted biological bacteria and host original flora

Abstract

The invention relates to the technical field of microbial preparation optimization and intelligent algorithm crossing, and particularly discloses a foraging algorithm for coupling dominant functional bacteria of targeted biological bacteria and host original flora, and the foraging algorithm simulates a coupled symbiotic system of the targeted biological bacteria and the host original flora into a flora society. Each to-be-optimized single-function bacterial strain is regarded as a'bacterial 'individual in the algorithm, the comprehensive effect of a flora on a specific target is used as a fitness function, global optimization is carried out through an improved chaos optimization bacterial foraging algorithm, and through the steps of flora proportion initialization, chemotaxis operation, breeding operation, migration operation and the like, the optimal bacterial foraging algorithm is obtained. The method provided by the invention solves the problems of complex work, long period and difficulty in obtaining a globally optimal solution in the traditional'trial-and-error method 'for preparing the composite flora, and provides core algorithm support for quickly, accurately and intelligently developing an efficient composite microbial preparation.

Description

Technical Field

[0001] This invention relates to the interdisciplinary field of computational biology, intelligent optimization algorithms and microbial technology, specifically to a foraging algorithm that couples dominant functional bacteria with the host's original microbial community. Background Technology

[0002] In agriculture and the environment, applying compound microbial agents to degrade pollutants and promote growth has become a green solution. Its core lies in coupling multiple functional strains (such as Bacillus and Lactobacillus) with the host's original microbial community to form a stable and efficient functional microbial community. However, determining the optimal combination and precise proportions of hundreds or thousands of strains is an extremely complex combinatorial optimization problem.

[0003] Currently, solving this problem mainly relies on the experience of researchers and a large number of trial-and-error experiments. This method has significant drawbacks: 1. Huge workload: It requires a massive amount of shake-flask culture and efficacy verification experiments, with a cycle that can last for months or even years.

[0004] 2. High cost: It consumes a lot of human, material and time resources.

[0005] 3. Difficulty in obtaining the optimal solution: Due to the huge search space, it is difficult to traverse all possible combinations based on experience, and it is easy to get trapped in local optima, making it impossible to obtain the globally optimal bacterial culture formula.

[0006] In the field of optimization algorithms, the Bacterial Foraging (BFO) algorithm is a swarm intelligence optimization algorithm that simulates the foraging behavior of E. coli. Due to its excellent global search capabilities, it is widely used in various engineering optimization problems. However, the standard BFO algorithm suffers from drawbacks such as slow convergence speed and susceptibility to local optima. Although some studies (such as CN107172627B) have combined it with chaos theory to optimize sensor network deployment, there are currently no reports of applying this algorithm across disciplines to the optimization of microbial community formulations in the life sciences.

[0007] Therefore, developing an algorithm that can intelligently and automatically search for optimal functional microbial community formulations has become an urgent need to promote the high-quality development of the microbial preparation industry. Summary of the Invention

[0008] The purpose of this invention is to provide a foraging algorithm that couples targeted biological bacteria with dominant functional bacteria of the host's original microbial community, aiming to solve the above-mentioned problems existing in the prior art.

[0009] The technical solution of the present invention to achieve the above objectives is as follows: The present invention provides a foraging algorithm for targeted microorganisms coupled with dominant functional microorganisms of the host's original microbiota, including the following steps: Step 1: Initialization configuration: Determine the specific types of candidate functional strains (such as growth-promoting bacteria, degradation bacteria, inactivating bacteria, etc.), and set the initial concentration range of each strain. Construct a multidimensional search space for the algorithm based on the strain types and concentration ranges; Set the core parameters of the algorithm, including chemotaxis number Nc, reproduction number Nre, migration number Ned, chemotaxis step size C(i), and basic migration probability Ped; Step 2: Chemotaxis: Simulate the functional fine-tuning process of the strain in the host environment. Use a random perturbation factor (perturbation amplitude is ±5%-±15% of the current bacterial population ratio) to perturb the current bacterial population ratio combination, and calculate the fitness value J(i,j,k) of the bacterial population combination for the preset target function after perturbation. Compare this fitness value with the fitness value before perturbation. If the fitness value increases by more than the preset threshold, continue to perform the "migratory" operation along the perturbation direction, and gradually optimize the bacterial population ratio according to the chemotactic step size C(i) until the fitness value no longer increases significantly. Step 3: Reproduction Operation: Simulate the survival of the fittest mechanism in natural selection, and calculate the cumulative fitness value of each bacterial community combination within a reproduction cycle (i.e., after completing Nc chemotaxis operations). The cumulative fitness value is the arithmetic mean of the fitness values ​​after each chemotaxis operation. Sort all bacterial community combinations according to the cumulative fitness value, eliminate the 50% of bacterial community combinations with the worst performance, and reproduce and divide the 50% of bacterial community combinations with the highest fitness value to ensure that the total population remains unchanged from the initial set number. Step 4: Migration Operation: Simulate the mutation effect in the natural environment. Based on the preset basic migration probability Ped, change the proportion distribution of some bacterial community combinations. By breaking the stability of the existing bacterial community proportions, the algorithm can escape local optima and explore new potential optimal bacterial community proportion combinations. Step 5: Termination condition judgment: preset maximum iteration value and optimal solution stability threshold; if the algorithm iteration count reaches the preset maximum iteration value, or the current optimal solution continuously reaches the stability threshold and remains unchanged, then the algorithm terminates and outputs the historical best bacterial community ratio combination; if none of the above termination conditions are met, return to step 2 and repeat the iteration process.

[0010] The foraging algorithm proposed in this invention, which couples targeted microorganisms with dominant functional microorganisms of the host's original microbial community, can efficiently screen out the optimal ratio of functional microorganisms that couple with the host's original microbial community by simulating the natural behaviors of microbial communities such as chemotaxis, reproduction, and migration, combined with algorithm iterative optimization. This solves the problems of traditional microbial community ratios relying on experience-based trial and error, poor coupling effect, and low target function achievement rate, and significantly improves the functional targeting and application effect of composite microbial communities.

[0011] Optionally, in the initialization configuration of step one, a chaotic mapping method is used to generate the initial bacterial community ratio. The specific implementation process is as follows: using the Logistic equation Z... n+1 =μZ n (1−Z n A chaotic sequence with a uniform distribution in the (0,1) interval is generated and mapped to the concentration range of each strain, thereby obtaining a more diverse initial bacterial community combination and improving the global search capability.

[0012] Optionally, in the chemotaxis operation of step two, a direction selection strategy based on chaotic perturbation is adopted. Specifically, when the bacterial community needs to perform a flipping operation, the completely random direction selection method is abandoned. Instead, a chaotic variable is generated through a Logistic chaotic sequence to guide the flipping direction. The direction vector is calculated using the formula: φ(i) = Δ(i) / , where Δ(i) is the direction vector guided by the chaotic sequence; by using the randomness and ergodicity of the chaotic variables, the search ergodicity of the chemotaxis operation is increased, and the probability of the algorithm getting trapped in local extrema is reduced.

[0013] Optionally, in the reproduction operation of step three, an elite retention and crossover mutation strategy is adopted, specifically: after sorting the bacterial populations by fitness, the top 50% of elite bacterial population combinations are retained; for the bottom 50% of bacterial population combinations, arithmetic crossover operation is performed between them and the elite bacterial populations to generate new progeny bacterial population combinations; and mutation operation is performed on some progeny bacterial populations to maintain population diversity.

[0014] Optionally, in step four, the migration probability is an adaptively dynamically adjusted probability. The specific adjustment process is as follows: before each migration operation, the fitness values ​​of all bacterial community combinations in the current population are calculated, and the maximum fitness value J of the current population is determined. health max Minimum fitness value J health min And the fitness value J of the individuals in the bacterial community currently to be adjusted. health According to the formula P self =(J health max -J health ) / (J health max -J health min )×Ped The actual migration probability of the individual is calculated; for fitness values ​​close to J... health min For individuals, the actual migration probability is close to the maximum value of Ped, while the fitness value is close to J. health maxFor individuals with poor performance, the actual migration probability is close to 0. This dynamic adjustment mechanism makes it easier for individuals with poor performance to migrate and mutate, while protecting the stability of high-performing individuals and balancing the algorithm's exploration ability and convergence speed.

[0015] Optionally, the fitness function J(i, j, k) is constructed according to the specific application goals of the complex microbial community, including but not limited to: - Passivation rate for heavy metals cadmium, lead, and arsenic; - Degradation rate of pesticides such as organophosphates and pyrethroids; - Efficiency in decomposing organic matter in the soil; - Contribution to crop growth promotion indicators; The fitness function can be a single metric or a weighted sum of the above-mentioned metrics.

[0016] This invention also provides a computer-readable storage medium storing a computer program thereon. The computer-readable storage medium includes, but is not limited to, media capable of storing computer programs such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, and optical disks. The computer program is stored in the storage medium in binary code form. When the computer program is executed by a processor, the execution modules of each step of the algorithm are called sequentially to implement all steps of the foraging algorithm for targeting bacteria and coupling dominant functional bacteria of the host's original flora as described in any of the above embodiments.

[0017] This invention also provides an intelligent formulation system for targeted biological agents, including a memory, a processor, an input module, an output module, and a communication module. The memory is a non-volatile memory used to store intermediate data generated during the execution of computer programs and algorithms. The processor is a central processing unit or a microcontroller unit used to call and execute the computer program stored in the memory, performing iterative optimization of the bacterial community ratio according to the algorithm steps. Specifically, this includes calling and executing an initialization module, a chemotactic calculation module, a propagation screening module, a migration adjustment module, and a termination judgment module. The input module includes a keyboard, a touchscreen, or a communication interface used to receive user-inputted algorithm parameters (such as Nc, Nre, Ped, etc.), functional strain information, concentration range, and application. The target (e.g., heavy metal passivation, pesticide degradation, etc.); the output module includes a display screen and a printer, used to display the algorithm's running status (e.g., iteration count, current optimal fitness value) and the final output optimal microbial community ratio combination; the communication module uses wired communication (e.g., RS485, Ethernet) or wireless communication (e.g., WiFi, Bluetooth) to establish a connection with the biological agent production equipment (e.g., strain fermentation tank, mixing and stirring equipment, quantitative feeding device); after the processor obtains the optimal microbial community ratio combination by executing the algorithm, it converts the ratio into the specific feeding amount of each functional strain (calculated in conjunction with the production scale), generates a standardized formulation plan, and sends it to the control unit of the production equipment through the communication module to guide the industrial-scale precise production and formulation of the targeted biological agent.

[0018] The beneficial effects of this invention are: 1. Intelligentization and Automation: Freeing formula design from tedious manual experiments, enabling intelligent optimization, significantly improving R&D efficiency, and shortening the R&D cycle.

[0019] 2. Strong global optimization capability: Combining the randomness and ergodicity of chaos theory with the various operational mechanisms of the BFO algorithm, it can effectively escape local optima and find the globally optimal or near-optimal bacterial community formula.

[0020] 3. Precise and efficient: Through mathematical modeling and calculation, it can quickly filter through a huge solution space, which is more precise and efficient than traditional methods.

[0021] 4. Versatility and Flexibility: By customizing the fitness function, this algorithm can be easily applied to the development of microbial agents with different objectives, such as degrading different pollutants and promoting the growth of different crops, demonstrating strong versatility. Detailed Implementation

[0022] The present invention will be further described below with reference to the embodiments. Example

[0023] Optimization Examples of Functional Microbial Community Formulations for Degrading Cadmium Pollution in Soil I. Problem Definition and Initialization: To address the issues of insufficient synergy between functional microbial agents and the original host microbial community in soil cadmium pollution remediation, and the unstable passivation effect, this embodiment aims to utilize the foraging algorithm described in the patent to screen the optimal concentration ratio of Bacillus subtilis, Bacillus mucilaginosus, Bacillus laterosporus, and the original host microbial community. This maximizes the passivation of available cadmium in the soil, ultimately improving the efficiency of soil cadmium pollution remediation and providing technical support for the safe utilization of farmland soil. Specifically, Bacillus subtilis, through its secreted extracellular polymers, can efficiently adsorb cadmium ions; Bacillus mucilaginosus can regulate soil pH to promote cadmium ion precipitation; Bacillus laterosporus can enhance the colonization stability of the microbial community in the soil environment; and the original host microbial community, as a native adapted microbial community, forms a coupled synergistic effect with the exogenous functional bacteria, jointly enhancing the cadmium passivation effect.

[0024] Taking into account the survival threshold, functional efficiency, and feasibility of industrial production of the strains in the soil environment, the concentration search range for all four bacterial groups was set at 1×10⁻⁶. 7 ~1×10 9 CFU / ml. This range ensures that the strain can form an effective number of colonies in the soil, avoiding functional failure due to excessively low concentration, while also avoiding resource waste and ecological risks caused by excessively high concentration.

[0025] The objective of this embodiment is to improve the cadmium passivation rate; therefore, the fitness function is directly set as the effective cadmium removal rate, i.e., J = Cd_removal_rate. The effective cadmium removal rate is determined by DTPA extraction-atomic absorption spectrophotometry.

[0026] Based on the microbial community optimization requirements in the soil cadmium passivation scenario, and after preliminary experimental verification, the key parameters of the algorithm are set as follows: bacterial population size S=50 (ensuring population diversity while controlling computational complexity); chemotaxis count Nc=100 (ensuring sufficient local search depth); reproduction count Nre=5 (balancing algorithm convergence speed and population diversity); migration count Ned=2 (avoiding excessive migration leading to the loss of high-quality formulas); basic migration probability Ped=0.25 (moderate initial exploration capability); chemotaxis step size C(i) is initially set to 5% of the concentration range (i.e., 5×10). 5 ~5×10 7 (CFU / ml) to ensure that the fine-tuning range can explore the adjacent optimization space without compromising the rationality of the bacterial community ratio.

[0027] Fifty initial bacterial community formulations were generated using the Logistic chaotic mapping method described in the patent. The specific steps are as follows: The Logistic equation control parameter μ=4 (ensuring the generated sequences are in a completely chaotic state), and a random initial value Zn (Zn∈(0,1) and Zn≠0.5 to avoid sequence degradation) was input. Fifty sets of chaotic sequences uniformly distributed in the (0,1) interval were generated iteratively using the formula Zn+1=4Zn(1−Zn). Subsequently, the linear mapping formula Cm=1×10⁻⁶ was used. 7 +Zn×(1×10 9 -1×10 7 The chaotic sequence was mapped to the concentration range of four strains, resulting in 50 initial bacterial community formulations with diverse distributions covering the entire search space, laying the foundation for the algorithm's global search.

[0028] II. Iterative Optimization: The algorithm performs iterative optimization according to the cyclical logic of chemotaxis, reproduction, and migration, with each operation closely linked, gradually screening out the optimal bacterial community formula: Chemotaxis: For 50 initial microbial community formulations, the algorithm simulates the functional fine-tuning process of strains in the soil environment. Each time, 1-2 strains are randomly selected, and their concentration ratios are slightly perturbed according to the initial chemotactic step size (5% of the concentration range). For example, in an initial formulation, the concentration of Bacillus subtilis is 3 × 10⁸ CFU / ml, and after perturbation, the concentration is adjusted to 2.85 × 10⁸ CFU / ml. 8 ~3.15×10 8 CFU / ml. After perturbation, the cadmium passivation rate of each formulation was rapidly determined using a 96-well plate miniaturized soil culture experiment: contaminated soil was sieved through a 2mm sieve, and the corresponding bacterial culture solution was added at a soil-to-water ratio of 1:2. The mixture was then incubated for 7 days at 25℃ and 60% field water holding capacity. The concentration of available cadmium was determined using DTPA extraction combined with atomic absorption spectrophotometry, and the fitness value J was calculated. If the passivation rate of a certain formulation after perturbation increased by more than 1% (preset threshold) compared to before perturbation, the adjustment direction was deemed effective, and the algorithm continued to iteratively adjust along this direction with a chemotactic step size until the passivation rate no longer significantly increased. If the passivation rate did not increase, a flip operation was triggered, and the bacterial community ratio was readjusted based on the direction vector guided by the Logistic chaotic sequence to ensure the ergonomic nature of the chemotactic operation.

[0029] Reproduction: After every 20 chemotaxis operations (i.e., one reproduction cycle), the algorithm calculates the cumulative fitness value of each bacterial culture formula within that cycle (the arithmetic mean of the passivation rate after 20 chemotaxis operations), thus comprehensively evaluating the stability of the formula. After sorting by cumulative fitness value in descending order, the 25 worst-performing formulas are eliminated, and the top 25 high-quality formulas are retained as elite bacterial cultures. To maintain the population size, the 25 vacant positions after elimination are filled with new formulas using an elite retention and crossover mutation strategy.

[0030] Migration: A migration operation is performed after each breeding operation. The adaptive migration probability mechanism described in the patent is used to dynamically adjust the migration probability of each formula based on the current population's fitness distribution: first, the maximum passivation rate J of the current population is calculated. health max and minimum passivation rate J health min Then through the formula P self =(J health max -J health ) / (J health max -J health min )×0.25 Calculate the actual migration probability for each formulation, for passivation rates close to J. health min The inferior formula, with an actual migration probability close to 0.25, had its bacterial community proportion randomly reset to a new value within the search space, introducing a completely new direction for exploration; for passivation rates close to J... health max The high-quality formula has an actual migration probability close to 0, ensuring the stability of the high-quality formula. Through this mechanism, the algorithm avoids getting trapped in local optima and protects the high-quality genes that have been screened.

[0031] III. Result Output: After 100 chemotaxis cycles, the algorithm converged. The historical best formula was output: Bacillus subtilis 5.8 × 10⁻⁶. 8 CFU / ml, Bacillus mucilaginosus 3.2×10 8 CFU / ml, Bacillus laterosporus brevis 2.1×10 8 CFU / ml, host original flora 1.5×10 8 CFU / ml. Full-scale validation experiments showed that this formulation achieved a 68.5% passivation rate for cadmium in soil, significantly higher than empirical formulations (average 45%) and other formulations in the optimization process.

[0032] in conclusion This invention successfully applies the chaotic optimization bacterial foraging algorithm to the optimization of functional microbial community formulations, demonstrating the effectiveness and superiority of the algorithm in this field, and providing a new technical path for high-throughput, intelligent microbial preparation development.

[0033] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They should not be construed as limiting the scope of protection of the present invention. All equivalent changes or modifications made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A foraging algorithm for targeted microorganisms coupled with dominant functional microorganisms of the host's original microbiota, characterized in that, Includes the following steps: Step 1: Initialization Configuration: Determine the types of strains to be selected and the initial concentration range of each strain, and construct the algorithm search space; The algorithm's core parameters are defined, including the number of chemotaxis Nc, the number of reproductions Nre, the number of migrations Ned, the chemotaxis step size C(i), and the basic migration probability Ped. Step 2: Chemotaxis: Simulate the process of fine-tuning the strain function, randomly perturb the current bacterial community ratio combination, and calculate the fitness value J(i,j,k) of the bacterial community combination for the preset target function after perturbation; if the fitness value is improved compared with that before perturbation, continue to perform the "migration" operation along the perturbation direction to optimize the bacterial community ratio; Step 3: Reproduction Operation: Simulating the natural selection mechanism, based on the cumulative fitness value of each bacterial community combination within a reproductive cycle, the 50% of bacterial community combinations with the worst performance are eliminated, and the 50% of bacterial community combinations with the highest fitness value are reproduced and divided to maintain the total population size unchanged; Step 4: Migration Operation: Simulate the effect of sudden environmental changes, alter the proportion distribution of some bacterial community combinations with a preset probability, and help the algorithm escape local optima and explore new bacterial community proportion combinations; Step 5: Termination condition judgment: If the number of algorithm iterations reaches the preset maximum iteration value, or the current optimal solution remains stable for a preset number of consecutive iterations, then terminate the algorithm and output the historical optimal bacterial community ratio combination; if the termination condition is not met, return to step (2) to repeat the iteration.

2. The foraging algorithm for targeted microorganisms coupled with dominant functional microorganisms of the host's original microbiota according to claim 1, characterized in that, In the initialization step one, a chaotic mapping method is used to generate the initial bacterial community proportions, specifically: using the Logistic equation Z... n+1 =μZ n (1−Z n A chaotic sequence with a uniform distribution in the (0,1) interval is generated and mapped to the concentration range of each strain, thereby obtaining a more diverse initial bacterial community combination and improving the global search capability.

3. The foraging algorithm for targeted microorganisms coupled with dominant functional microorganisms of the host's original microbiota according to claim 1, characterized in that, In the chemotaxis operation of step two, a direction selection strategy based on chaotic perturbation is adopted. Specifically, when the bacterial community needs to perform a flipping operation, the completely random direction selection method is abandoned. Instead, a chaotic variable is generated through a Logistic chaotic sequence to guide the flipping direction. The direction vector is calculated using the formula: φ(i) = Δ(i) / , where Δ(i) is the direction vector guided by the chaotic sequence.

4. The foraging algorithm for targeted microorganisms coupled with dominant functional bacteria of the host's original microbiota according to claim 1, characterized in that, In the reproduction operation of step three, an elite retention and crossover mutation strategy is adopted. Specifically, after sorting the bacterial populations by fitness, the top 50% of elite bacterial population combinations are retained; for the bottom 50% of bacterial population combinations, arithmetic crossover operation is performed with the elite bacterial populations to generate new progeny bacterial population combinations; and mutation operation is performed on some progeny bacterial populations to maintain population diversity.

5. The foraging algorithm for targeted microorganisms coupled with dominant functional bacteria of the host's original microbiota according to claim 1, characterized in that, In the migration operation of step four, the migration probability Ped is adaptive and dynamically adjusted according to the fitness of individual microbial individuals. Individuals with poor fitness have a higher migration probability, as shown in the formula: P self =(J health max -J health ) / (J health max -J health min )×Ped, Among them, J health max J health min J represents the maximum and minimum fitness values ​​of the current population, respectively. health This represents the fitness value of the current individual.

6. A foraging algorithm for targeted microorganisms coupled with dominant functional microorganisms of the host's original microbiota according to any one of claims 1-5, characterized in that, The fitness function J(i, j, k) is constructed according to the specific application goals of the complex microbial community, including but not limited to: - Passivation rate for heavy metals cadmium, lead, and arsenic; - Degradation rate of pesticides such as organophosphates and pyrethroids; - Efficiency in decomposing organic matter in the soil; - Contribution to crop growth promotion indicators; The fitness function can be a single metric or a weighted sum of the above-mentioned metrics.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the algorithm as described in any one of claims 1-6.

8. An intelligent formulation system for targeted biological agents, characterized in that, It includes a memory and a processor, the memory storing a computer program, and the processor executing the program to implement the steps of the algorithm as described in any one of claims 1-6, and to guide the production and formulation of the biological agent based on the optimal bacterial community ratio output by the algorithm.