A smart identification method and system for maize-nitrogen-fixing bacteria interactions

Abstract

This invention relates to the field of bioinformatics analysis technology, specifically to an intelligent identification method and system for maize-nitrogen-fixing bacteria interactions. The method includes the following steps: obtaining encoded information and secretion concentrations from maize rhizosphere samples; classifying secretion trends to generate root behavior response data; screening bacteria with consistent behavior to construct a microbial community structure sequence; identifying and outputting interaction pattern labels based on convergent combinations; ranking and stabilizing labels and adjusting the order of output fields to generate a sorted structure for the behavior label fields. This invention enhances the ability to identify the synergistic characteristics of rhizosphere microbial community behavior by introducing behavior trend classification and bacterial species sorting identification. Combined with a label selection mechanism based on co-occurrence frequency and trend structure, it achieves consistent sorting and semantic aggregation of the label output content, optimizes the accuracy of label structure expression under diverse interaction patterns, and improves the structural stability and information controllability of the output fields by using a frequency distribution-driven field priority adjustment method.

Description

Technical Field

[0001] This invention relates to the field of bioinformatics analysis technology, and in particular to an intelligent identification method and system for maize-nitrogen-fixing bacteria interactions. Background Technology

[0002] The field of bioinformatics analysis technology encompasses computer-aided modeling, analysis, and identification of genetic information, biomolecules, and related interaction processes within organisms. Its core content involves acquiring data, extracting features, and modeling systems from genetic material, protein expression, metabolites, and interaction networks in biological samples using digital methods. This results in a data processing system that can be used for biological function prediction, germplasm screening, disease analysis, or environmental response research. Bioinformatics analysis covers tasks such as gene sequencing, protein structure analysis, biological community behavior, microbial interactions, and the identification and modeling of various complex life processes. It relies on the coordinated advancement of multiple directions, including algorithm design, biological data preprocessing, information fusion, and knowledge graph construction.

[0003] The intelligent identification method for maize-nitrogen-fixing bacteria interactions refers to an analytical process that involves acquiring metagenomic data from the coexisting environment of maize and rhizosphere microorganisms, identifying bacterial species, functional proteins, and expressed genes related to nitrogen fixation, and annotating and discriminating their interactions with the maize host. This process includes nucleic acid extraction based on biological samples, obtaining community metagenomic data through high-throughput sequencing, screening potential nitrogen-fixation-related protein genes using protein sequence alignment algorithms, constructing a maize-microorganism interaction network based on a protein interaction database, and identifying the interaction relationships between specific species or strains and the maize host using sample classification and labeling methods. This involves metagenomic data acquisition, feature protein screening, biological interaction annotation, and data comparison and identification.

[0004] Traditional maize nitrogen-fixing bacteria interaction identification technology suffers from problems in practical applications, including coarse-scale bioinformatics data collection and insufficient precision in interaction relationship identification. It fails to fully utilize the changing trends of plant exudates and microbial activity in the maize rhizosphere environment, resulting in a lack of sufficient and accurate temporal and spatial correlation information in interaction relationship identification. This affects the effectiveness of microbial-host behavioral correlation analysis. Protein interaction network construction relies on static protein data comparison, ignoring dynamic microbial behavioral characteristics and their changing trends over different time periods. This makes it difficult to accurately classify the interaction patterns between bacterial species. When the interactions of bacterial communities are complex or involve multiple response modes, it is difficult to clearly distinguish the specific behavioral types of different bacterial combinations. Consequently, the output interaction tags have a certain degree of ambiguity in practical biological research or agricultural production applications, limiting the effective utilization of interaction information and restricting its application effectiveness. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the existing technology and to propose an intelligent identification method and system for maize-nitrogen-fixing bacteria interaction.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a smart identification method for maize-nitrogen-fixing bacteria interactions, comprising the following steps:

[0007] S1: Using maize rhizosphere samples, obtain the plant number, sampling date, layer depth identifier, spatial orientation number and sampling batch number for each sample, detect the concentration values ​​of organic acids, sugars and phenols, and construct a sample information list;

[0008] S2: Call the sample information list, analyze the changing trends of sugar, organic acid and phenol release in the rhizosphere region at multiple time points and classify them by direction, screen samples with significant changes by matching with the corresponding sample time groups, establish a hierarchical mapping relationship for continuous trend segments, and generate root behavior response data.

[0009] S3: Call the root behavior response data, analyze the direction of change in the activity intensity of the bacterial species in multiple time periods, compare it with the root behavior characteristics, screen the bacterial species with the same direction, classify them according to the frequency of recurrence and the magnitude of behavior, and generate a bacterial community behavior structure sequence.

[0010] S4: Analyze the behavioral structure sequence of the bacterial community, identify the co-occurrence of each bacterial species combination in multiple rhizosphere samples, analyze and identify the convergence of behavioral performance of each combination in multiple time periods, identify the relationship type based on the behavioral similarity between bacterial communities, output interaction pattern labels, and generate bacterial community interaction pattern information.

[0011] As a further embodiment of the present invention, the sample information list includes location information parameters, secretion concentration parameters, and number index fields; the root behavior response data includes behavior trend direction, time period level label, and change range; the microbial community behavior structure sequence includes microbial species combination identifier, activity change label, and behavior frequency distribution; and the microbial community interaction mode information specifically includes interaction type label, behavior synergy relationship, and combination response characteristics.

[0012] As a further aspect of the present invention, the step of obtaining the sample information list specifically includes:

[0013] S111: Using maize rhizosphere samples, obtain the plant number, sampling date, layer depth identifier, spatial orientation number and sampling batch number for each sample. Through format unification, field alignment and structural classification, construct a standardized coded data set and generate a basic sample number sequence.

[0014] S112: Call the sample basic number sequence to detect the concentration of organic acids, sugars and phenols in the sample, assign the detection data according to the sampling batch and spatial location, record the data and locate the concentration values ​​in combination with the sample order, and generate a rhizosphere exudate concentration dataset.

[0015] S113: Based on the rhizosphere exudate concentration dataset, according to the sample number field and concentration index, the sample data is arranged sequentially according to the sampling time and layer depth position, the sample data is structured and organized into an indexable list, and a sample information list structure unit is generated.

[0016] As a further aspect of the present invention, the step of obtaining the root behavior response data specifically includes:

[0017] S211: Call the sample information list, extract the concentration detection data of organic acids, sugars and phenols, analyze the release amount of each type of secretion at multiple consecutive time points, analyze the consistency of the release direction at adjacent time points, classify the direction in combination with the numerical trend, and generate a secretion trend classification result set.

[0018] S212: Based on the secretion trend classification result set, extract the mapping data between the classification direction and the time group, compare the release trend changes of multiple samples, calculate the trend aggregation index of each sample, analyze the differences in change magnitude, type response and trend consistency, and screen the samples to generate sample trend screening distribution results.

[0019] S213: Call the sample trend filtering distribution results, sequentially number each sample type, establish a hierarchical mapping relationship for continuous trend segments, and generate root behavior response data.

[0020] As a further aspect of the present invention, the step of obtaining the bacterial community behavior structure sequence specifically includes:

[0021] S311: Call the root behavior response data to obtain the activity intensity sequence of each bacterial species in multiple time periods, analyze the change direction at each time point, establish a direction classification record, and generate a bacterial species behavior direction sequence;

[0022] S312: Based on the bacterial species behavior direction sequence, compare the behavior type and direction marker in the root behavior response data of each bacterial species, calculate the behavior consistency strength coefficient of the bacterial species combination, screen bacterial species with consistent directions, and generate a bacterial community behavior trend index.

[0023] S313: Based on the community behavior trend index, group the microbial community according to the frequency of recurrence in each time period, classify the time distribution range and behavioral performance amplitude of the grouping results, and generate a microbial community behavior structure sequence.

[0024] As a further aspect of the present invention, the step of obtaining the microbial community interaction mode information specifically includes:

[0025] S411: Obtain the behavioral structure sequence of the microbial community, analyze the number of times each group of microbial species combinations appears simultaneously in the rhizosphere sample, calculate the co-occurrence frequency value, and generate the co-occurrence frequency value of the microbial species combination;

[0026] S412: Call the co-occurrence frequency value of the bacterial species combination, calculate the consistency of the behavioral change direction of the bacterial species in the combination over multiple time periods, calculate the behavioral convergence coefficient of the bacterial species combination, and generate the behavioral convergence level of the bacterial species combination;

[0027] S413: Call the behavioral convergence level of the bacterial species combination, and based on the behavioral similarity between bacterial communities, identify the relationship type through cluster analysis, including synchronous response, cooperative action, one-way influence, and mutual interference and repulsion, output interaction pattern labels, and generate bacterial community interaction pattern information.

[0028] As a further aspect of the present invention, the method further includes:

[0029] S5: Call the community interaction mode information, analyze the sorting structure of the bacterial behavior trend in each type of interaction mode, filter the label items with the same sorting position, extract the bacterial identification number, behavior trend label and root response level number, sort the priority according to the frequency distribution of behavior trend, adjust the sorting order of label output fields, and generate the behavior label field sorting structure.

[0030] The behavior tag field sorting structure includes field arrangement order, priority mapping tags, and response attribute tags.

[0031] As a further aspect of the present invention, the step of obtaining the sorting structure of the behavior tag field specifically includes:

[0032] S511: Call the bacterial community interaction mode information, analyze the sorting structure of bacterial species behavior trends in each interaction mode, filter and identify the label items that maintain a consistent sorting position in multiple interaction modes, and generate a stable sorting label item set.

[0033] S512: Based on the stable sorted tag item set, extract the corresponding strain identification number, behavior trend tag and root response level number, sort them according to the frequency distribution of behavior trends, and generate a tag item priority order table.

[0034] S513: Call the tag item priority order table, adjust the order of tag output fields, and generate a sorting structure for behavior tag fields.

[0035] A smart identification system for maize-nitrogen-fixing bacteria interactions, the system being used to execute the aforementioned smart identification method for maize-nitrogen-fixing bacteria interactions, the system comprising:

[0036] The rhizosphere sampling module collects plant number, sampling date, layer depth identifier, spatial orientation number and batch information based on maize rhizosphere samples. Combined with the detection of organic acid, sugar and phenolic substance concentrations, it establishes a sample information list.

[0037] Based on the sample information list, the behavior trend mapping module analyzes the release trends of organic acids, sugars and phenols at multiple time nodes. By matching directional classification with time grouping, it filters samples with significant changes, establishes a hierarchical mapping relationship for continuous trend segments, and generates root behavior response data.

[0038] Based on the root behavior response data, the microbial strain screening module compares and analyzes the direction of change in microbial activity intensity with the direction of rhizosphere secretion trend, screens microbial strains whose activity trend is consistent with the direction of rhizosphere behavior, and establishes a microbial community behavior structure sequence according to the frequency of repetition and the magnitude of behavioral performance.

[0039] The relationship pattern clustering module performs co-occurrence and convergence analysis of bacterial species combinations based on the bacterial community behavior structure sequence, evaluates the behavioral similarity between multiple combinations, outputs interaction pattern labels, and generates bacterial community interaction pattern information.

[0040] The tag structure rearrangement module analyzes the sorting structure of bacterial behavior trends in each type of interaction mode based on the bacterial community interaction mode information, filters tag items with consistent sorting positions, extracts bacterial identifiers, behavior trends and root response levels, prioritizes and sorts the behavior trends according to their frequency of occurrence, adjusts the order of tag output fields, and establishes a sorting structure for behavior tag fields.

[0041] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0042] In this invention, by introducing behavioral trend classification and species sorting identification, the ability to identify the collaborative characteristics of rhizosphere microbial community behavior is enhanced. By combining the label item screening mechanism based on co-occurrence frequency and trend structure, the sorting consistency and semantic aggregation of the label output content are achieved, optimizing the expression accuracy of the label structure under the diversity interaction mode. By using the field priority adjustment method driven by frequency distribution, the structural stability and information controllability of the output field are improved. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of the workflow of the present invention;

[0044] Figure 2 This is a flowchart of the sample information list acquisition process of the present invention;

[0045] Figure 3 This is a flowchart of the root behavior response data acquisition process of the present invention;

[0046] Figure 4 This is a flowchart of the process for obtaining the microbial community behavior structure sequence of the present invention;

[0047] Figure 5 This is a flowchart of the process for obtaining microbial community interaction mode information in this invention.

[0048] Figure 6 This is a flowchart illustrating the process of obtaining the sorting structure for the behavior tag field in this invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0050] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0051] Please see Figure 1 This invention provides a technical solution: an intelligent identification method for maize-nitrogen-fixing bacteria interactions, comprising the following steps:

[0052] S1: Using maize rhizosphere samples, obtain the plant number, sampling date, layer depth identifier, spatial orientation number and sampling batch number for each sample, detect the concentration values ​​of organic acids, sugars and phenols, and construct a sample information list;

[0053] S2: Call the sample information list, analyze the changing trends of sugar, organic acid and phenol release in the rhizosphere region at multiple time points and classify the directions, screen samples with significant changes by matching with the corresponding sample time groups, establish a hierarchical mapping relationship for continuous trend segments, and generate root behavior response data.

[0054] S3: Call root behavior response data, analyze the direction of change in the activity intensity of bacterial species over multiple time periods, compare it with root behavior characteristics, screen bacterial species with the same direction, classify them according to the frequency of recurrence and the magnitude of behavior, and generate a bacterial community behavior structure sequence.

[0055] S4: Analyze the behavioral structure sequence of the microbial community, identify the co-occurrence of each group of microbial species in multiple rhizosphere samples, analyze and identify the convergence of behavioral performance of each combination in multiple time periods, identify the relationship type based on the behavioral similarity between microbial communities, output interaction pattern labels, and generate microbial community interaction pattern information.

[0056] S5: Call the microbial community interaction mode information, analyze the sorting structure of microbial behavior trends in each type of interaction mode, filter the label items with the same sorting position, extract the microbial identifier number, behavior trend label and root response level number, sort them according to the frequency distribution of behavior trends, adjust the sorting order of label output fields, and generate the behavior label field sorting structure.

[0057] The sample information list includes location information parameters, secretion concentration parameters, and number index fields. The root behavioral response data includes behavioral trend direction, time period level labels, and change range. The microbial community behavioral structure sequence includes microbial combination identifiers, activity change labels, and behavioral frequency distribution. The microbial community interaction mode information specifically includes interaction type labels, behavioral synergy relationships, and combination response characteristics. The behavioral label field sorting structure includes field arrangement order, priority mapping labels, and response attribute labels.

[0058] Please see Figure 2 The specific steps for obtaining the sample information list are as follows:

[0059] S111: Using maize rhizosphere samples, obtain the plant number, sampling date, layer depth identifier, spatial orientation number and sampling batch number for each sample. Through format unification, field alignment and structural classification, construct a standardized coded data set and generate a basic sample number sequence.

[0060] Based on the actual rhizosphere sampling points deployed within the maize planting area, each sample collection unit was initially labeled according to the field number to which the plant belonged. The sampling time was recorded on-site and converted to a YYYY-MM-DD format for timestamp identification, assigning a sampling date identifier. The root system was divided into layers according to depths of 0-20cm, 20-40cm, and 40-60cm, with corresponding identification codes A, B, and C. For each sample, the sampling direction was assigned directional numbers 1, 2, 3, and 4 in a clockwise order based on the four cardinal directions (north, south, east, west). Batch numbers were determined by the experimental sequence, uniformly set as "P + three-digit serial number," e.g., the first sampling batch number was P001, the second P002, etc. A field set was established by sequentially calling the above five types of fields, and the original sampling data structure was formatted uniformly, with all fields categorized according to "plant number." The data is concatenated and combined in the order of "Number-Date-Depth Identifier-Direction Number-Batch Number". Sample data with missing fields are deleted. Then, a two-dimensional array structure is built with the field order as the column index. A duplicate detection process is performed on each sample number item. If duplicates are found, the data at the last time point is used as the main data and the preceding data is deleted. After constructing the number matrix, the samples are grouped and statistically analyzed to ensure that the data of different batches and different direction numbers are completely covered. For example, when the plant with number C015 completes the data recording on April 12, 2025, at depth B, direction number 3, and batch P007, the basic sample number is "C015-2025-04-12-B-3-P007". After all the standardized numbers are completed, the number list is output and an index mapping table is automatically built according to the number order. This process finally yields the basic sample number sequence.

[0061] S112: Call the basic sample number sequence to detect the concentration of organic acids, sugars and phenols in the sample. Identify the detection data according to the sampling batch and spatial location. Combine the sample order to record data and locate concentration values ​​to generate a rhizosphere exudate concentration dataset.

[0062] The system reads the pre-constructed sample base number sequence and, through program control, sequentially executes concentration detection commands on the samples in the number list. For each sample, the sensor module is invoked to collect concentrations of organic acids, sugars, and phenols. The data unit is set to μg / g. Sampling batch numbers P001-P010 correspond to a total of 10 sampling batches. Each sampling direction is a four-way coverage. In the example, the sample number "B023-2025-04-15-A-1-P005" shows organic acid concentrations of 55.3 μg / g and sugar concentrations of 84.7 μg / g. Phenolic compounds were collected at 22.6 μg / g. After collection, a two-level index table was constructed by separating the direction number field and the batch number field according to the sample number. All sampled samples were grouped according to their batches to create a first-level index, and each batch was further grouped according to the direction number to create a second-level index. During the data writing process, the index path was automatically matched according to the basic sample number, and the concentration value was written into the structured table. At the same time, the row and column positions of the concentration values ​​in the data structure were recorded. The concentration field was set as the column name, and the sample number index was set as the row name. Finally, a complete rhizosphere exudate concentration dataset was generated. The detection results of some samples are shown in Table 1.

[0063] Table 1. Data on root exudate concentration

[0064] Sample number Organic acids (μg / g) Carbohydrates (μg / g) Phenolic compounds (μg / g) C015-2025-04-12-B-3-P007 49.2 77.6 28.9 B023-2025-04-15-A-1-P005 55.3 84.7 22.6 D041-2025-04-18-C-4-P006 60.1 69.3 31.4

[0065] As shown in Table 1, the concentration values ​​of the three types of secretions in different samples have been associated with the test results through standardized numbering and the data have been entered into the table.

[0066] S113: Based on the rhizosphere exudate concentration dataset, the sample data is arranged in order of sampling time and layer depth according to the sample number field and concentration index, and the sample data is structured and organized into an indexable list to generate sample information list structure units.

[0067] The above rhizosphere exudate concentration dataset underwent structured processing. First, the date field was extracted from the sample numbers. All samples were sorted in ascending chronological order. The layer depth identifiers A, B, and C from the sample numbers, corresponding to the 0-20cm, 20-40cm, and 40-60cm layers respectively, were used as vertical layer indices. During structure construction, the date field was set as the primary time axis index, and the layer depth identifiers as the secondary axis index. The direction and batch numbers in the sample numbers were excluded, retaining only the spatial and temporal dual indices needed to construct the behavioral trend. After matrix reorganization, the three types of concentration data were categorized into their respective fields. A three-dimensional structure list was constructed for each field, with the structural dimensions being time, layer depth, and sample number. In this structure, the corresponding number, such as "C015-2025-04-12-B-3-P007", will be positioned in the three matrices of rhizosphere exudate concentration, recording its organic acid concentration of 49.2 μg / g, sugar concentration of 77.6 μg / g, and phenol concentration of 28.9 μg / g. The index position is 2025-04-12, layer depth B, and number C015. An indexable field list is established through programming logic to achieve matching records between fields and concentration values, so that subsequent operations can directly retrieve the corresponding concentration index values ​​according to the date and layer depth dimensions. After the entire structuring and indexing process is completed, a complete information unit with number hierarchy, concentration field, and concentration value is output, generating a sample information list structure unit.

[0068] Please see Figure 3 The specific steps for obtaining root behavior response data are as follows:

[0069] S211: Call the sample information list, extract the concentration detection data of organic acids, sugars and phenols, analyze the release amount of each type of secretion at multiple consecutive time points, analyze the consistency of the release direction at adjacent time points, combine the numerical trend to classify the direction, and generate a secretion trend classification result set.

[0070] The system calls the structural unit of the sample information list to extract the concentration data of three types of secretions—organic acids, sugars, and phenols—from each sample record. Taking the actual corn plant number C015 as an example, the concentration data from five consecutive time points between April 12, 2025 and April 16, 2025 were monitored to create a concentration value array. The array order is arranged chronologically. For example, the organic acid concentrations of sample C015 were 49.2, 52.1, 48.5, 55.3, and 57.4 μg / g; the sugar concentrations were 77.6, 80.2, 78.1, 81.4, and 84.7 μg / g; and the phenolic concentrations were 28.9, 30.4, 27.5, 31.1, and 33.2 μg / g. For each secretion, the difference in concentration values ​​between two consecutive days was calculated to determine the release direction. Specifically, the difference between the concentration values ​​on April 12, 2025 and April 16, 2025 was calculated. The concentration change of organic acids was 52.1-49.2 = +2.9 μg / g, defined as an "increase" in release direction. From April 13th to April 14th, 2025, the change was 48.5-52.1 = -3.6 μg / g, defined as a "decrease" in release direction. Subsequent changes in direction were calculated and recorded sequentially. The results of changes in direction at multiple consecutive time points were compared item by item. Specifically, it was determined whether the direction between consecutive adjacent points was consistent. Consecutive points with the same direction were counted as one trend unit, and those with different directions were marked as trend reversal. The number of trend units was determined for the three types of secretions. The trend units for organic acids were "increase-decrease-increase", for sugars it was "increase-decrease-increase", and for phenols it was "increase-decrease-increase". The trend units of the three types of substances in the sample were classified according to the trend direction. After the classification was completed, a summary table of trend units was output, which formed the secretion trend classification result set.

[0071] S212: Based on the secretion trend classification result set, extract the mapping data between classification direction and time group, compare the release trend changes of multiple samples, and use the formula:

[0072]

[0073] Calculate the trend aggregation index for each sample, analyze the differences in change magnitude, category response, and trend consistency, and filter the samples to generate sample trend filtering distribution results;

[0074] Among them, P i Z represents the trend convergence index of sample i. ik This represents the normalized concentration value of the k-th secretion in sample i. L represents the normalized average concentration of all secretions from sample i. i C represents the number of times a trend reversal occurs in sample i. i D represents the number of time periods corresponding to sample i. iThe number of other samples with the same trend direction as sample i is indicated by n, where n represents the number of secretion types involved in the calculation in this sample, k represents the secretion type number, and i represents the sample number.

[0075] Based on the result set of secretion trend classification, the trend direction and corresponding time point data of each sample are extracted one by one. Taking sample C015 as an example, Z is determined. ik The value represents the normalized concentration of the k-th secretion in the sample over the entire time period. The calculation steps are: subtract the minimum concentration of each type from the actual concentration, then divide by the difference between the maximum and minimum concentrations of that type. Taking organic acids as an example, among their five concentration values ​​of 49.2, 52.1, 48.5, 55.3, and 57.4 μg / g, the maximum is 57.4 μg / g and the minimum is 48.5 μg / g. The first normalized concentration value is... Five normalized concentration values ​​were obtained sequentially, and their mean was calculated. For example, if the mean of the samples (0.082, 0.377, 0.000, 0.755, 1.000) is 0.443, calculate the absolute difference between the normalized concentration value and the mean for each type of secretion. Then, sum and average the results for all samples to obtain the first result. Subsequently, count the number of trend reversals, L. i For example, if the trend reversal occurred twice in sample C015, then the number of time periods C... i The number D of other samples with the same direction i To perform statistical analysis, for example, sample C015 corresponds to 4 time periods, and there are a total of 6 samples with the same trend direction. Substituting these values ​​into the formula:

[0076]

[0077] Where n = 3, L i =2, C i =4, D i =6, the actual calculation is:

[0078]

[0079] The calculation process is as follows:

[0080]

[0081] The trend convergence index is a scoring indicator used to measure the comprehensive volatility and typicality of a single sample in the trend of root exudate changes. It reflects whether the sample's behavior is stable and representative. The smaller the parameter value, the more concentrated the release trend of various exudates in the sample, the more consistent the direction of change, and the closer it is to the trend pattern of other samples in the overall sample, making it a typical sample. The larger the index, the higher the dispersion and frequency of the release behavior of the sample, the greater the difference in trend performance from other samples, constituting a special abnormal response group. The index is used for downstream screening analysis, identifying key response areas, and supporting sample screening and label structure matching tasks in subsequent interaction modeling. The results show that a high trend convergence index (an index exceeding 1.0 is considered high) indicates significant trend changes. Based on this, the convergence index was calculated and sorted for other samples one by one, and samples with an index exceeding 0.8 were selected to form the sample trend screening distribution results.

[0082] S213: Call the sample trend filtering distribution results, sequentially number each sample type, establish the hierarchical mapping relationship of continuous trend segments, and generate root behavior response data;

[0083] The sample trend filtering distribution results are retrieved, and samples that meet the aggregation index filtering criteria are numbered sequentially according to their index values, from largest to smallest. The sample with the highest aggregation index is designated as T1, the second highest as T2, and so on. The number of trend segments corresponding to each number is recorded, and a level mapping is performed based on the number of trend segments. One to two trend segments represent a first-level trend, three to four segments represent a second-level trend, and five or more segments represent a third-level trend, as shown in Table 2.

[0084] Table 2 Sample Trend Level Mapping Table

[0085] Sample number Aggregation Index number of trend paragraphs Trend Level T1(C015) 1.212 3 Secondary trend T2(B023) 0.975 4 Secondary trend T3(D041) 0.881 2 Primary Trend

[0086] As shown in Table 2, sample C015 has 3 trend segments and is classified as a secondary trend. Similarly, other samples are numbered one by one and their hierarchical mapping relationships are determined. Finally, root behavior response data is generated.

[0087] Please see Figure 4 The specific steps for obtaining the microbial community behavioral structure sequence are as follows:

[0088] S311: Call the root behavior response data to obtain the activity intensity sequence of each species in multiple time periods, analyze the change direction at each time point, establish a direction classification record, and generate a species behavior direction sequence;

[0089] The structural units in the root behavior response data are invoked to extract and process the activity intensity of each bacterial species within the corresponding time period. Taking the bacterial species numbered L07 as an example, its original activity intensity values ​​in five consecutive time periods are 42.1, 47.6, 44.2, 49.3, and 45.5 μmol / g, respectively. To calculate the directional change of this sequence, the difference between consecutive time points must first be determined. From April 10th to April 11th, 2025, the difference is 47.6 - 42.1 = +5.5, indicating an "enhancing" direction. From April 11th to April 12th, 2025, the difference is 44.2 μmol / g. -47.6 = -3.4, which is judged as "weakening". From April 12 to April 13, 2025, it is 49.3 - 44.2 = +5.1, which is "enhancing". From April 13 to April 14, 2025, it is 45.5 - 49.3 = -3.8, which is "weakening". Therefore, the activity direction sequence of strain L07 is "enhancing-weakening-enhancing-weakening". After numerical labeling with "enhancing" as 1 and "weakening" as -1, a direction array {1,-1,1,-1} is formed. The behavioral change direction of all sample strains is recorded using the same rule to generate a strain behavior direction sequence matrix.

[0090] S312: Based on the bacterial species' behavioral direction sequence, compare the behavioral type and the direction markers in the root behavior response data for each bacterial species using the following formula:

[0091]

[0092] Calculate the behavioral consistency strength coefficient of bacterial species combinations, screen bacterial species with consistent orientations, and generate bacterial community behavior trend indicators;

[0093] Where l represents the strain number, t represents the time period number, m is the total number of time periods, b is the total number of strains screened, j is the strain number, and F l,t R represents the normalized value of the activity intensity of bacterial strain l at time t. t M represents the normalized value of the root behavior's response intensity at time point t. l L represents the normalized value of the time coverage range of bacterial strain l. l A represents the normalized value of the expression variation range of bacterial species l. j To screen the frequency of the behavioral direction of strain j, B l H is the normalized value of the fluctuation range of the behavioral intensity of strain l. l The coefficient representing the behavioral consistency of the bacterial ensemble;

[0094] Based on the bacterial behavior sequence obtained in the previous step, each bacterial species was numbered. Taking strain L07 as an example, its activity intensity at each time point was normalized. The original activity value sequence was set to 42.1, 47.6, 44.2, 49.3, and 45.5 μmol / g, with a maximum value of 49.3 and a minimum value of 42.1. After normalization, the sequence was 0.000, 0.764, 0.294, 1.000, and 0.490, defined as F. l,t The normalized response intensity R of the root behavior response data at the same time point t Given values ​​of 0.15, 0.42, 0.39, 0.67, and 0.51, calculate the first term. Right now:

[0095] Set the normalized value M for the time coverage range of this bacterial strain. l =0.89, representing the normalized value L of the variation range. l =0.65, then The difference between the two items is |1.356 - 1.102| = 0.254. Let B be the fluctuation range of the bacterial species' behavior intensity. l =0.73. A total of 5 bacterial strains participated in this consistency screening. The average frequency of behavioral direction for each strain was calculated. The final consistency strength coefficient is:

[0096]

[0097] The behavioral consistency strength coefficient of the microbial ensemble is an index reflecting the consistency of microbial behavior with the rhizosphere environment throughout the observation period. It integrates the interaction between the activity intensity and rhizosphere behavioral response intensity of the microbial species across multiple time periods, considering both the temporal variation and the stability of behavioral intensity. This quantifies the behavioral consistency of the microbial ensemble, indicating the synchronicity of microbial behavior with the rhizosphere environment and the stability of the microbial species' own behavior. It helps assess the interaction between the microbial species and the rhizosphere environment, as well as the behavioral consistency of the microbial community. A higher coefficient indicates more stable behavior within a given time period and a more consistent interaction with the rhizosphere environment, while a lower coefficient suggests greater fluctuations in microbial behavior over time or a weaker synergistic effect with rhizosphere behavior. This coefficient allows for comparisons among different microbial species, identifying which species exhibit strong synergistic effects in specific rhizosphere environments, thus providing a basis for subsequent identification of microbial community interaction patterns. The results show that the final behavioral consistency strength coefficient for species L07 is 0.0129.

[0098] S313: Based on the community behavior trend indicators, group the microbial community according to the frequency of recurrence in each time period, classify the time distribution range and behavioral performance amplitude of the grouping results, and generate a microbial community behavior structure sequence.

[0099] After obtaining the behavioral consistency intensity index for all bacterial species, a frequency matrix was established based on whether each species repeatedly appeared in the consistency screening results across different time periods. With the time period as the horizontal axis and the species number as the vertical axis, the matrix recorded whether each species was selected in each time period. Species appearing more than 3 times were grouped together, and their time coverage was further mapped to the difference between the start and end numbers of the time periods. For example, L07 was screened 4 times across 5 time periods, with a start time of 2025-04-10 and an end time of 2025-04-14, covering a time span of 5 days, and was classified as high coverage. Frequency-based classification was then performed, followed by calculation of behavioral amplitude differences based on the maximum activity difference of the group of bacteria within the time period. The maximum value of L07 was 49.3 μmol / g, the minimum value was 42.1 μmol / g, and the amplitude was 7.2 μmol / g. The behavioral amplitude classification threshold was set at 5.0 μmol / g, so this group was classified as a high amplitude group. Finally, based on the classification logic of "high frequency + high amplitude", this type of bacteria was grouped into primary bacterial community behavioral structure units. The same operation was performed on the remaining bacteria in turn, and a complete bacterial community behavioral structure sequence was generated according to the three-dimensional indicators of frequency, amplitude, and time span.

[0100] Please see Figure 5 The specific steps for obtaining information on microbial community interaction patterns are as follows:

[0101] S411: Obtain the microbial community behavior structure sequence, analyze the number of times each microbial species combination appears simultaneously in the rhizosphere sample, calculate the co-occurrence frequency value, and generate the microbial species combination co-occurrence frequency value;

[0102] The generated microbial community behavior structure sequence is called, and each group of microbial species combinations is extracted and their co-occurrence in each rhizosphere sample number is statistically analyzed. The sample number range is selected from C015 to C024, with a total of 10 samples. Microbial species combinations present in the samples, such as combination P03, which includes species L07, L11, and L15, are selected. Boolean judgments are performed on whether these three species appear simultaneously in each sample. If all three species exist in the corresponding behavior structure of a sample, the sample is counted as one co-occurrence. The sample number to which it belongs is recorded and the co-occurrence count is accumulated. For example, if combination P03 exists simultaneously in samples C015, C017, C019, and C021, the co-occurrence frequency value is counted as 4. The co-occurrence value is mapped to the combination number to construct a key-value mapping structure from combination number to co-occurrence frequency. Finally, the co-occurrence frequency of each microbial species combination is calculated, as shown in the example below.

[0103] Table 3. Co-occurrence frequency of bacterial species combinations

[0104] Combination number Includes bacterial strains (number). Co-occurrence frequency value Co-occurring sample number P01 L03, L06, L10 3 C015, C016, C022 P02 L01, L05, L09 5 C016, C017, C019, C020, C023 P03 L07, L11, L15 4 C015, C017, C019, C021

[0105] As shown in Table 3, combination P02 has the highest co-occurrence frequency value, and further analysis needs to measure the convergence of its behavioral performance.

[0106] S412: Calculate the consistency of behavioral changes in the combined bacterial strains over multiple time periods by calling the co-occurrence frequency values, using the following formula:

[0107]

[0108] Calculate the behavioral convergence coefficient of the strain combination and generate the behavioral convergence level of the strain combination;

[0109] Among them, G p M represents the behavioral convergence coefficient of the p-th bacterial species combination. pq This represents the intensity value of the behavior of the p-th group in the q-th time period. S represents the average intensity of behavioral performance across all time periods for combination p. pq Y represents the co-occurrence frequency of the p-th combination in the q-th time period. p Z represents the number of time periods in combination p. p X represents the number of combinations with similar behavioral patterns to combination p. p Q represents the standard deviation of the performance intensity of combination p in different time periods, where p represents the combination number and Q represents the total number of time periods.

[0110] Using the co-occurrence frequency values ​​listed in Table 3 and combining them with the behavioral intensity data of each group of bacteria in each time period, taking combination P03 as an example, its behavioral intensity in the five time periods was 9.3, 10.2, 8.5, 11.6, and 10.0 μmol / g, respectively, denoted as M. pq The average behavioral intensity of combination P03 over these five time periods was calculated as follows: The intensity at each time point was subtracted from the mean, and the absolute value was taken as 0.62, 0.28, 1.42, 1.68, and 0.08, respectively. These values ​​were then multiplied by the corresponding co-occurrence frequency value S. pq If the co-occurrence frequencies of this combination in the five time periods are 3, 2, 2, 3, and 1 respectively, then:

[0111]

[0112] The number of time periods Y in combination P03 p =5, Number of similar behavior combinations Z p =3, standard deviation X p Calculated from the standard deviation of behavioral intensity, the standard deviation calculation formula is as follows:

[0113]

[0114] Then the behavioral convergence coefficient G p for:

[0115]

[0116] Among them, the behavioral convergence coefficient of bacterial ensembles is a numerical indicator used to quantitatively describe the consistency and concentration of the common behaviors of bacterial communities in samples across multiple time periods. This parameter comprehensively measures the volatility of the behavioral intensity of each group of bacteria at different time periods, the frequency structure of their co-occurrence in the samples, and the representativeness of the ensemble within the system. It reflects whether the ensemble belongs to a categorizable and stably expressed microbial interaction pattern. A smaller value indicates a more concentrated, stable, and reliable co-occurrence behavior; a larger value indicates an unstable behavioral pattern, and the interaction may be sporadic or disruptive. This parameter drives the labeling and classification of ensemble relationship types and interaction pattern structures, and is a core bridging parameter in the interaction inference chain. According to the established criteria, the convergence levels are divided as follows: G p <2.5 indicates low convergence, 2.5≤G p <4.0 indicates moderate convergence, G p A value of ≥4.0 indicates high convergence, therefore the convergence level of the combination P03 is "high".

[0117] S413: Call the behavioral convergence level of the bacterial species combination, identify the relationship type through cluster analysis based on the behavioral similarity between bacterial communities, including synchronous response, cooperative action, one-way influence, mutual interference and repulsion, output interaction pattern labels, and generate bacterial community interaction pattern information;

[0118] The similarity assessment of behavioral characteristics of all bacterial species combinations is performed by calling the behavioral convergence level of each combination. The behavioral time curves are matched and clustered. Combinations with high convergence levels and similar co-occurrence frequencies in the same time period are grouped into one category. If the behavioral directions of two combinations are synchronized and the time delay does not exceed one time period, they are identified as "synchronous response". If the two combinations have the same direction in more than three time periods and the numerical difference is less than 20%, they are classified as "cooperative action". If the behavioral trend of combination A always appears one time period after combination B, it is labeled as "unidirectional influence". If the behavioral directions of combination A and B are opposite in most time periods and the intensity difference exceeds 30%, it is labeled as "mutual interference and repulsion". Based on this logic, P01 and P02 are classified into the "cooperative action" category, and P02 and P03 are classified into the "unidirectional influence" category. A combination interaction mode label matrix is ​​established. The output result is the bacterial community interaction mode information, which includes three items: combination number, mode type and behavioral description structure.

[0119] Please see Figure 6The specific steps for obtaining the sorting structure of the behavior tag field are as follows:

[0120] S511: Call the microbial community interaction mode information, analyze the sorting structure of the microbial behavior trend in each interaction mode, filter and identify the label items that maintain a consistent sorting position in multiple interaction modes, and generate a stable sorting label item set.

[0121] The microbial community interaction mode information structure unit is invoked to extract the specific ranking structure of the microbial behavior trend under each interaction mode. The analysis is conducted using the mode types "cooperative action" and "one-way influence" as examples. The mode numbers involved in the analysis are set as K01 (cooperative action) and K02 (one-way influence). For the microbial combination P01 (L03→L06→L10) and combination P02 (L01→L05→L09) in mode K01, the ranking numbers are recorded as 1, 2, and 3 respectively. Similarly, for the combination P02 (L01→L05→L09) and P03 (L07→L11→L15) in mode K02, the ranking number is also recorded as 1. 2, 3. Then, a cross-comparison analysis was performed on the sorting numbers of all bacterial combinations. Specifically, the corresponding positions of the sorting numbers of P01 and P02, and P02 and P03 were compared in different modes. The sorting position of combination P02 was L01 1st, L05 2nd, and L09 3rd in both modes. It was determined that the sorting structure of this combination was a stable sorting. However, the sorting positions of the bacterial species in combination P01 and P03 were not consistent in different modes. They were determined to be unstable sorting. Finally, the stable sorting label items in all modes were recorded to obtain a set of stable sorting label items. The example set contains bacterial species L01, L05, and L09 of combination P02.

[0122] S512: Based on a stable set of sorted label items, extract the corresponding strain identification number, behavior trend label and root response level number, sort them according to the frequency distribution of behavior trends, and generate a label item priority order table.

[0123] Based on the aforementioned stable set of sorted labels, the strain IDs (L01, L05, L09) are extracted item by item from the set. The corresponding behavioral trend labels and root response level numbers are then retrieved. Taking L01 as an example, the behavioral trend labels are "continuous enhancement," "continuous enhancement," and "enhancement after weakening," and the root response level numbers are 2, 2, and 1, respectively. The frequency of the labels is then counted. The "continuous enhancement" label appears most frequently, counted twice, and the "enhancement after weakening" label appears once. Based on the frequency distribution statistics, the label priorities are explicitly sorted, i.e., frequency. The highest priority label, "Continuous Enhancement," is assigned the highest priority number 1, and the second highest frequency label, "Enhancement After Weakening," is assigned the second-lowest priority number 2. Subsequently, the labels within the same frequency are sorted and adjusted according to the root response level number. For example, labels with higher root response levels (larger numbers) are sorted first, and those with lower levels are sorted last. For example, the root response level number corresponding to L01 in the "Continuous Enhancement" label is 2, so it is prioritized over the label item with level number 1. The above steps are performed on all stable sorted bacterial species labels in this way to confirm them one by one, and finally a clear label priority order table is formed, as shown in Table 4.

[0124] Table 4. Priority Order of Tag Items

[0125] Priority number strain number Behavioral trend tags Root response level number 1 L01 Continuous enhancement 2 2 L05 Continuous enhancement 1 3 L09 weakened then strengthened 1

[0126] As shown in Table 4, strain L01 has the highest priority, and the priority ranking results have been clearly completed.

[0127] S513: Call the tag item priority order table, adjust the order of tag output fields, and generate the behavior tag field sorting structure;

[0128] Call the tag item priority order table in Table 4 and perform the tag field output order adjustment operation. Use the priority number field as the basis for the tag output order and redefine the field arrangement position. Specifically, place the strain number with the smallest priority number and its corresponding behavior trend label and root response level number field at the beginning of the output data structure. For example, in the L01 strain item in Table 4, its priority number is the smallest (number 1), and the corresponding data is output first, arranged in the order of "strain number L01 → behavior trend label continuously enhanced → root response level number 2". Then, according to the same rule, arrange the corresponding data of strain L05 with number 2 and strain L09 with number 3 in the subsequent positions, output in the order of "L05 → continuously enhanced → 1" and "L09 → enhanced after weakening → 1", to complete the complete adjustment of the field position and form the final behavior label field sorting structure used for output.

[0129] A smart identification system for maize-nitrogen-fixing bacteria interactions, the system being used to execute the aforementioned smart identification method for maize-nitrogen-fixing bacteria interactions, the system comprising:

[0130] The rhizosphere sampling module collects plant number, sampling date, layer depth identifier, spatial orientation number and batch information based on maize rhizosphere samples. Combined with the detection of organic acid, sugar and phenolic substance concentrations, it establishes a sample information list.

[0131] The behavioral trend mapping module analyzes the release trends of organic acids, sugars and phenols at multiple time points based on the sample information list. By matching directional classification with time grouping, it screens samples with significant changes, establishes a hierarchical mapping relationship for continuous trend segments, and generates root behavioral response data.

[0132] The microbial strain screening module compares the direction of change in microbial activity intensity with the direction of rhizosphere secretion trend based on root behavior response data. It screens microbial strains whose activity trend is consistent with the direction of rhizosphere behavior and establishes a microbial community behavior structure sequence according to the frequency of repetition and the magnitude of behavioral performance.

[0133] The relationship pattern clustering module performs co-occurrence and convergence analysis of bacterial species combinations based on the bacterial community behavior structure sequence, evaluates the behavioral similarity between multiple combinations, outputs interaction pattern labels, and generates bacterial community interaction pattern information.

[0134] The label structure reordering module analyzes the sorting structure of bacterial behavior trends in each type of interaction mode based on the information of bacterial community interaction patterns, filters the label items with the same sorting position, extracts the bacterial species identifier, behavior trend and root response level, sorts them according to the frequency of occurrence of behavior trends, adjusts the sorting order of label output fields, and establishes the sorting structure of behavior label fields.

[0135] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A smart identification method for maize-nitrogen-fixing bacteria interactions, characterized in that, Includes the following steps: S1: Using maize rhizosphere samples, obtain the plant number, sampling date, layer depth identifier, spatial orientation number and sampling batch number for each sample, detect the concentration values ​​of organic acids, sugars and phenols, and construct a sample information list; S2: Call the sample information list, analyze the changing trends of sugar, organic acid and phenol release in the rhizosphere region at multiple time points and classify them by direction, screen samples with significant changes by matching with the corresponding sample time groups, establish a hierarchical mapping relationship for continuous trend segments, and generate root behavior response data. The specific steps for obtaining the root behavior response data are as follows: S211: Call the sample information list, extract the concentration detection data of organic acids, sugars and phenols, analyze the release amount of each type of secretion at multiple consecutive time points, analyze the consistency of the release direction at adjacent time points, classify the direction in combination with the numerical trend, and generate a secretion trend classification result set. S212: Based on the secretion trend classification result set, extract the mapping data between classification direction and time group, compare the release trend changes of multiple samples, and use the formula: ； Calculate the trend aggregation index for each sample, analyze the differences in change magnitude, category response, and trend consistency, and filter the samples to generate sample trend filtering distribution results; in, Indicates sample The trend convergence index Indicates sample No. Normalized concentration values ​​of the secretions Indicates sample Normalized average concentration of all secretions Indicates sample The number of times a trend reversal occurs. Indicates sample The number of corresponding time periods Indicates trend direction and sample The same number of other samples, This indicates the number of secretion types included in the calculation for this sample. Indicates the type number of the secretion. Indicates the sample number; S213: Call the sample trend filtering distribution results, sequentially number each sample type, establish a hierarchical mapping relationship for continuous trend segments, and generate root behavior response data; S3: Call the root behavior response data, analyze the direction of change in the activity intensity of the bacterial species in multiple time periods, compare it with the root behavior characteristics, screen the bacterial species with the same direction, classify them according to the frequency of recurrence and the magnitude of behavior, and generate a bacterial community behavior structure sequence. The specific steps for obtaining the microbial community behavioral structure sequence are as follows: S311: Call the root behavior response data to obtain the activity intensity sequence of each bacterial species in multiple time periods, analyze the change direction at each time point, establish a direction classification record, and generate a bacterial species behavior direction sequence; S312: Based on the bacterial species' behavioral direction sequence, compare the behavioral type and the direction markers in the root behavior response data for each bacterial species using the following formula: ； Calculate the behavioral consistency strength coefficient of bacterial species combinations, screen bacterial species with consistent orientations, and generate bacterial community behavior trend indicators; in, Represents the strain number, Represents a time period number. The total number of time periods. The total number of bacterial strains screened. Number the bacterial strain. Representative strains In time The normalized value of activity intensity Represents the root behavior at a given time point. The normalized value of the response intensity, Representative strains Normalized value of time coverage Representative strains The normalized value of the range of expression variation. To screen bacterial strains Frequency of behavioral direction strains The normalized value of the amplitude of behavioral intensity fluctuation. The coefficient representing the behavioral consistency of the bacterial ensemble; S313: Based on the community behavior trend index, group the microbial community according to the frequency of recurrence in each time period, classify the time distribution range and behavioral performance amplitude of the grouping results, and generate a microbial community behavior structure sequence. S4: Analyze the behavioral structure sequence of the microbial community, identify the co-occurrence of each group of microbial species in multiple rhizosphere samples, analyze and identify the convergence of behavioral performance of each combination in multiple time periods, identify the relationship type based on the behavioral similarity between microbial communities, output interaction pattern labels, and generate microbial community interaction pattern information. The specific steps for obtaining the microbial community interaction pattern information are as follows: S411: Obtain the behavioral structure sequence of the microbial community, analyze the number of times each group of microbial species combinations appears simultaneously in the rhizosphere sample, calculate the co-occurrence frequency value, and generate the co-occurrence frequency value of the microbial species combination; S412: Calculate the consistency of behavioral changes in the combined bacterial strains over multiple time periods by calling the co-occurrence frequency values, using the following formula: ； Calculate the behavioral convergence coefficient of the strain combination and generate the behavioral convergence level of the strain combination; in, Indicates the first Behavioral convergence coefficient of individual bacterial species combinations Indicates the first The combination in the first Intensity values ​​of behavioral performance over a given time period Indicate combination The average intensity of behavioral performance over all time periods. Indicates the first The combination in the first Co-occurrence frequency values ​​for each time period Indicate combination Number of time periods Representation and Combination The number of combinations with similar behavioral patterns Indicate combination The standard deviation of performance intensity over different time periods The number indicating the combination of bacterial strains. Indicates the total number of time periods; S413: Call the behavioral convergence level of the bacterial species combination, and based on the behavioral similarity between bacterial communities, identify the relationship type through cluster analysis, including synchronous response, cooperative action, one-way influence, mutual interference and repulsion, output interaction mode labels, and generate bacterial community interaction mode information; S5: Call the microbial community interaction mode information, analyze the sorting structure of microbial behavior trends in each type of interaction mode, filter the label items with the same sorting position, extract the microbial identifier number, behavior trend label and root response level number, sort them according to the frequency distribution of behavior trends, adjust the order of label output fields, and generate the behavior label field sorting structure.

2. The intelligent identification method for maize-nitrogen-fixing bacteria interaction according to claim 1, characterized in that, The sample information list includes location information parameters, secretion concentration parameters, and number index fields. The root behavior response data includes behavior trend direction, time period level label, and change range. The microbial community behavior structure sequence includes microbial combination identifier, activity change label, and behavior frequency distribution. The microbial community interaction mode information specifically includes interaction type label, behavior synergy relationship, and combination response characteristics. The behavior label field sorting structure includes field arrangement order, priority mapping label, and response attribute label.

3. The intelligent identification method for maize-nitrogen-fixing bacteria interaction according to claim 1, characterized in that, The specific steps for obtaining the sample information list are as follows: S111: Using maize rhizosphere samples, obtain the plant number, sampling date, layer depth identifier, spatial orientation number and sampling batch number for each sample. Through format unification, field alignment and structural classification, construct a standardized coded data set and generate a basic sample number sequence. S112: Call the sample basic number sequence to detect the concentration of organic acids, sugars and phenols in the sample, assign the detection data according to the sampling batch and spatial location, record the data and locate the concentration values ​​in combination with the sample order, and generate a rhizosphere exudate concentration dataset. S113: Based on the rhizosphere exudate concentration dataset, according to the sample number field and concentration index, the sample data is arranged sequentially according to the sampling time and layer depth position, the sample data is structured and organized into an indexable list, and a sample information list structure unit is generated.

4. The intelligent identification method for maize-nitrogen-fixing bacteria interaction according to claim 1, characterized in that, The specific steps for obtaining the sorting structure of the behavior label field are as follows: S511: Call the bacterial community interaction mode information, analyze the sorting structure of bacterial species behavior trends in each interaction mode, filter and identify the label items that maintain a consistent sorting position in multiple interaction modes, and generate a stable sorting label item set. S512: Based on the stable sorted tag item set, extract the corresponding strain identification number, behavior trend tag and root response level number, sort them according to the frequency distribution of behavior trends, and generate a tag item priority order table. S513: Call the tag item priority order table, adjust the order of tag output fields, and generate a sorting structure for behavior tag fields.

5. A smart identification system for maize-nitrogen-fixing bacteria interactions, characterized in that, The system is used to implement the intelligent identification method for maize-nitrogen-fixing bacteria interaction as described in any one of claims 1-4, and the system comprises: The rhizosphere sampling module collects plant number, sampling date, layer depth identifier, spatial orientation number and batch information based on maize rhizosphere samples. Combined with the detection of organic acid, sugar and phenolic substance concentrations, it establishes a sample information list. Based on the sample information list, the behavior trend mapping module analyzes the release trends of organic acids, sugars and phenols at multiple time nodes. By matching directional classification with time grouping, it filters samples with significant changes, establishes a hierarchical mapping relationship for continuous trend segments, and generates root behavior response data. Based on the root behavior response data, the microbial strain screening module compares and analyzes the direction of change in microbial activity intensity with the direction of rhizosphere secretion trend, screens microbial strains whose activity trend is consistent with the direction of rhizosphere behavior, and establishes a microbial community behavior structure sequence according to the frequency of repetition and the magnitude of behavioral performance. The relationship pattern clustering module performs co-occurrence and convergence analysis of bacterial species combinations based on the bacterial community behavior structure sequence, evaluates the behavioral similarity between multiple combinations, outputs interaction pattern labels, and generates bacterial community interaction pattern information. The tag structure rearrangement module analyzes the sorting structure of bacterial behavior trends in each type of interaction mode based on the bacterial community interaction mode information, filters tag items with consistent sorting positions, extracts bacterial identifiers, behavior trends and root response levels, prioritizes and sorts the behavior trends according to their frequency of occurrence, adjusts the order of tag output fields, and establishes a sorting structure for behavior tag fields.

Images