Peanut anti-bacterial wilt breeding strategy analysis method and system based on growth cycle

Abstract

The invention relates to the technical field of data processing, and discloses a peanut anti-bacterial wilt breeding strategy analysis method and system based on a growth cycle. The method comprises the following steps: acquiring and analyzing peanut growth period characteristic data to obtain a growth period characteristic library; performing phenotype comparison on healthy plants and diseased plants, and establishing an anti-bacterial wilt early warning model; performing real-time monitoring according to the early warning model to obtain a risk early warning result; performing growth period tracking based on an early warning result, and establishing a resistance scoring system; strain stability parameters are obtained through environmental response analysis; and finally determining excellent strains to form a bacterial wilt resistant breeding scheme. By establishing a data monitoring and analysis system in the whole growth period, dynamic evaluation and early warning of the bacterial wilt resistance of the plant are realized, and the accuracy and efficiency of bacterial wilt resistance variety breeding are improved.

Description

Technical Field

[0001] This application relates to the field of data processing, and particularly to a method and system for analyzing peanut breeding strategies for bacterial wilt resistance based on the growth cycle. Background Art

[0002] Peanut is an important oil and cash crop. Cultivated peanuts are self-pollinated crops with a relatively narrow genetic basis. The existing germplasm resources and traditional breeding techniques are difficult to further improve the breeding efficiency. Bacterial wilt, as a bacterial disease, is currently a major disease of peanuts, mainly concentrated in the central and southern regions and showing a trend of gradually spreading northward.

[0003] The existing peanut breeding methods for bacterial wilt resistance mainly rely on field observation and empirical judgment, lacking systematic analysis and utilization of growth cycle data. In the traditional breeding process, the evaluation of bacterial wilt resistance is often limited to the phenotypic observation at a specific period, making it difficult to comprehensively reflect the bacterial wilt resistance performance of plants throughout the growth period. At the same time, environmental factors have a greater impact on the evaluation of bacterial wilt resistance, resulting in insufficient stability and reliability of the screening results. In addition, the existing technologies also have problems such as single evaluation index, insufficient data analysis, and lack of early warning mechanism, which all restrict the improvement of the breeding efficiency for bacterial wilt resistance. Summary of the Invention

[0004] This application provides a method and system for analyzing peanut breeding strategies for bacterial wilt resistance based on the growth cycle, which realizes the dynamic evaluation and early warning of plant bacterial wilt resistance by establishing a data monitoring and analysis system for the entire growth period, and improves the accuracy and efficiency of breeding bacterial wilt-resistant varieties.

[0005] In a first aspect, the present application provides a method for analyzing peanut bacterial wilt resistance breeding strategies based on the growth cycle. The method for analyzing peanut bacterial wilt resistance breeding strategies based on the growth cycle includes: continuously collecting the morphological characteristics, physiological indicators, and disease symptoms of peanuts during the growth period through a phenotypic monitoring system, dividing the data according to the key growth stages through time series segmentation, constructing a time series feature vector, and obtaining a growth cycle feature library; based on the growth cycle feature library, comparing and analyzing the phenotypic trajectories of healthy plants and diseased plants, establishing early warning indicators through differential feature extraction, and obtaining a bacterial wilt resistance early warning model; according to the bacterial wilt resistance early warning model, monitoring and analyzing the growth phenotypic data of breeding materials in real time, determining potential susceptible individuals through warning threshold judgment, and obtaining a risk warning result; based on the risk warning result, tracking the whole growth period of plants with different resistance levels, establishing a resistance scoring system through comparison of phenotypic characteristics at key stages, and obtaining a dynamic resistance evaluation plan; according to the dynamic resistance evaluation plan, collecting multi-point test data for environmental response analysis, establishing adaptability indicators through plant-environment interaction evaluation, and obtaining strain stability parameters; according to the strain stability parameters, comprehensively evaluating and grading breeding materials, and determining excellent strains through a grading and screening system, so as to obtain a peanut bacterial wilt resistance breeding plan based on the growth cycle.

[0006] In a second aspect, the present application provides a system for analyzing peanut bacterial wilt resistance breeding strategies based on the growth cycle. The system for analyzing peanut bacterial wilt resistance breeding strategies based on the growth cycle includes: A collection module, which is used to continuously collect the morphological characteristics, physiological indicators, and disease symptoms of peanuts during the growth period through a phenotypic monitoring system, divide the data according to the key growth stages through time series segmentation, construct a time series feature vector, and obtain a growth cycle feature library; A comparison module, which is used to compare and analyze the phenotypic trajectories of healthy plants and diseased plants based on the growth cycle feature library, establish early warning indicators through differential feature extraction, and obtain a bacterial wilt resistance early warning model; A monitoring module, which is used to monitor and analyze the growth phenotypic data of breeding materials in real time according to the bacterial wilt resistance early warning model, determine potential susceptible individuals through warning threshold judgment, and obtain a risk warning result; A tracking module, which is used to track the whole growth period of plants with different resistance levels based on the risk warning result, establish a resistance scoring system through comparison of phenotypic characteristics at key stages, and obtain a dynamic resistance evaluation plan; An analysis module, which is used to collect multi-point test data for environmental response analysis according to the dynamic resistance evaluation plan, establish adaptability indicators through plant-environment interaction evaluation, and obtain strain stability parameters; A grading module is used to comprehensively evaluate and grade breeding materials according to the strain stability parameters, determine excellent plant lines through a grading and screening system, and obtain a breeding plan for bacterial wilt resistance based on the growth cycle. In the technical solution provided by this application, by continuously collecting and segmenting time series of the morphological characteristics, physiological indicators, and disease symptoms during the peanut growth period, a complete growth cycle feature library is constructed, realizing comprehensive monitoring and recording of the plant growth and development process. Through comparative analysis of the phenotypic trajectories of healthy plants and diseased plants, an early warning index system is established, providing data support for timely discovery of potential susceptible individuals. A deep learning algorithm is introduced into the bacterial wilt resistance early warning model, and using the time series analysis ability of a recurrent neural network, the growth phenotypic data is monitored and analyzed in real time, improving the accuracy of early warning. By tracking the entire growth period of plants with different resistance levels and comparing the phenotypic characteristics at key periods, a dynamic resistance evaluation plan is established, making the evaluation of bacterial wilt resistance more objective and comprehensive. Through environmental response analysis of multi-point test data and plant-environment interaction evaluation, an adaptability index system is constructed, enhancing the stability of strain selection. Finally, a comprehensive evaluation of breeding materials is carried out through a grading and screening system, realizing the precise screening of excellent plant lines. This method applies artificial intelligence algorithms to the field of peanut bacterial wilt resistance breeding. By using a long short-term memory network to extract features and perform time series analysis on plant growth data, the bacterial wilt resistance information contained in the growth cycle data is fully mined, providing a data-driven scientific decision-making basis for the breeding of bacterial wilt-resistant varieties. At the same time, this method establishes a complete data processing and analysis process, realizing the full-process intelligence from data collection, feature extraction, early warning analysis to strain screening, greatly improving the efficiency and accuracy of the bacterial wilt resistance breeding work. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0008] Figure 1 It is a schematic diagram of an embodiment of the method for analyzing the breeding strategy of peanut bacterial wilt resistance based on the growth cycle in the embodiments of this application; Figure 2 It is a schematic diagram of the hybridization and mating process in the embodiments of this application; Figure 3 It is a schematic diagram of the process of dividing data according to key growth periods through time series segmentation, constructing time series feature vectors, and obtaining a growth cycle feature library in the embodiments of this application; Figure 4This is a schematic diagram of an embodiment of the peanut bacterial wilt resistance breeding strategy analysis system based on the growth cycle in the embodiments of the present application. Detailed implementation manners

[0009] The embodiments of the present application provide a peanut bacterial wilt resistance breeding strategy analysis method and system based on the growth cycle. Terms such as "first", "second", "third", "fourth", etc. (if any) in the specification, claims and the above-mentioned drawings of the present application are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "comprising" or "having" and any deformation thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0010] For ease of understanding, the specific process of the embodiments of the present application will be described below. Please refer to Figure 1 An embodiment of the peanut bacterial wilt resistance breeding strategy analysis method based on the growth cycle in the embodiments of the present application includes: Step S101: Continuously collect the morphological characteristics, physiological indicators and disease symptoms of peanuts during the growth period through a phenotype monitoring system, divide the data according to the key growth stages through time series segmentation, construct a time series feature vector, and obtain a growth cycle feature library; Step S102: According to the growth cycle feature library, compare and analyze the phenotype trajectories of healthy plants and diseased plants, and establish early warning indicators through differential feature extraction to obtain a bacterial wilt resistance early warning model; Step S103: Based on the bacterial wilt resistance early warning model, monitor and analyze the growth phenotype data of breeding materials in real time, and determine potential susceptible individuals through the judgment of the early warning threshold to obtain a risk warning result; Step S104: Based on the risk warning result, track the whole growth period of plants with different resistance levels, and establish a resistance scoring system through the comparison of phenotypic characteristics at key stages to obtain a dynamic resistance evaluation plan; Step S105: According to the dynamic resistance evaluation plan, collect multi-point test data for environmental response analysis, and establish adaptability indicators through plant-environment interaction evaluation to obtain strain stability parameters; Step S106: Based on the strain stability parameters, comprehensively evaluate and classify the breeding materials, and determine excellent strains through a grading and screening system to obtain a bacterial wilt resistance breeding plan based on the growth cycle.

[0011] It can be understood that the execution entity of this application can be an analysis system for peanut breeding strategies against bacterial wilt based on the growth cycle, or it can also be a terminal or a server. Specifically, it is not limited here. In the embodiments of this application, the server is taken as the execution entity for illustration.

[0012] Specifically, as Figure 2 shown, it is a schematic diagram of the hybridization and breeding process in the embodiments of this application. Among them: The female parent selects mutagenic materials with strong comprehensive stress resistance and resistance to bacterial wilt, and the male parent selects high-yield and high-quality varieties. F1-generation individuals are obtained through hybridization. After collecting and analyzing the growth period data of F1-generation individuals, self-crossing is carried out to obtain the F2-generation population. The F2-generation is planted in a bacterial wilt nursery for resistance identification, and the selected excellent individual plants enter the F3-generation for southern multiplication and generation addition. The F3-generation materials are successively subjected to line comparison tests and strain comparison tests, and then the stability of the strains is evaluated through multi-point tests. Finally, the excellent strains enter the regional test, and excellent strains with stable resistance to bacterial wilt are selected through comprehensive evaluation. In the stage of selecting breeding materials, a peanut variety with strong comprehensive stress resistance and resistance to bacterial wilt is selected as the female parent, and a high-yield and high-quality peanut variety is selected as the male parent for hybridization and grouping. After obtaining the F1-generation hybrid individuals, the growth period data of the F1-generation individuals are collected. The data collection of the peanut growth period is divided into three key index dimensions: the morphological characteristics dimension collects plant height, number of branches, and leaf area parameters; the physiological index dimension collects photosynthesis intensity, respiration intensity, and chlorophyll content; the disease symptom dimension collects lesion area, disease progression, and incidence data. The data collection runs through three key growth stages: the seedling stage, the vegetative growth stage, and the reproductive growth stage. The collected original data is segmented according to different growth stages in a time series, and the data of each growth period is constructed into a feature matrix. The rows in the matrix represent different plant individuals, and the columns represent various monitoring indicators. The feature matrix is subjected to dimensionality reduction processing to extract the main features to form a time series feature vector. The elements in the time series feature vector are arranged in chronological order, reflecting the growth and development and the performance characteristics of resistance to bacterial wilt of the plants during the entire growth cycle. These time series feature vectors are summarized to form a growth cycle feature library.

[0013] Based on the growth cycle feature library of F1-generation individuals, the plants are divided into a healthy group and an infected group according to their health status. The morphological characteristics, physiological indicators, and disease symptom data of the two groups of plants are analyzed by time series comparison, and the feature difference values at each time point between the two groups are calculated. The calculation of the difference value uses the standardized numerical difference to eliminate the influence of the dimension between different indicators. The difference values are arranged in a time series to construct a phenotypic trajectory difference matrix. Each row in the matrix represents a time point, and each column represents a feature index. Based on the difference matrix, statistical analysis is carried out to calculate the difference significance of each feature in different growth stages, and the key index combinations that can give early warnings are screened out to establish a warning model for resistance to bacterial wilt.

[0014] During the F2 generation breeding process, the population obtained by self-crossing the F1 generation was planted in a bacterial wilt disease nursery for resistance identification. At the same time, the F2 generation population was monitored in real time according to the bacterial wilt resistance early warning model. The growth phenotype data of each plant was input into the early warning model, and risk assessment was carried out through the set early warning threshold. The setting of the early warning threshold was based on the statistical differences in the characteristics of healthy plants and diseased plants in the early stage, and the index threshold with the highest significance difference was selected. When the phenotypic characteristics of the plant exceeded the early warning threshold, it was marked as a potentially susceptible individual. In this way, risk early warning was carried out on the F2 generation population to timely detect individuals with poor resistance performance. For plants with different resistance levels selected from the F2 generation, directional breeding of the F3 generation was continued. During the process of accelerating generations in the south of China for the F3 generation, plants with different resistance levels were tracked throughout the growth period, and the phenotypic characteristic data at each key growth stage were recorded. A phenotypic characteristic scoring standard was established, and different characteristic indicators were quantified into scores to form a resistance scoring system. The formulation of the scoring standard was based on the numerical range of characteristic values, and continuous data was converted into discrete scores. In this way, the resistance level of the plants was quantitatively evaluated to obtain a dynamic resistance evaluation plan.

[0015] The excellent strain lines were put into the strain comparison test stage, planted at multiple test sites, and environmental factor data such as soil physical and chemical properties and climate conditions at each test site were collected. The environmental factor data was correlated with the plant phenotypic data to establish the plant-environment interaction relationship. The genotype effect, environmental effect, and interaction effect were separated by variance decomposition to evaluate the environmental adaptability of the strain lines and calculate the stability parameters. Materials with excellent stability parameter performance were selected to enter the regional test stage. A comprehensive evaluation system including stability indicators, bacterial wilt resistance indicators, and yield indicators was established to score the test materials in multiple dimensions. The scoring results were classified, and a classification and screening standard was established. Based on the classification standard, strain lines with excellent comprehensive performance were selected to formulate a bacterial wilt resistance breeding plan.

[0016] For example, in the collection of morphological characteristics of F1 hybrid materials, the plant height of each individual plant was measured, and the initial data included continuous measurement values at 30 time points during the growth period. These time points were segmented according to the growth stages. The seedling stage, vegetative growth stage, and reproductive growth stage each included 10 measurement points. The mean and coefficient of variation were calculated for the data of each period to form period characteristic values. At the same time, physiological indicators such as photosynthetic intensity and chlorophyll content, as well as disease indicators such as lesion size, were recorded. These indicators were arranged in chronological order to construct a time series data matrix containing multiple feature dimensions. The data matrix was dimensionally reduced to extract the main features to form feature vectors. In subsequent early warning analysis, the health status of the plants was judged by calculating the similarity between these feature vectors and the standard feature vectors in the early warning model. Based on this data analysis method, a batch of individual plants with excellent resistance performance was successfully screened out in the F2 population and further verified in the F3 generation. Through the analysis of multi-point test data, the stability of these excellent individual plants was evaluated, and excellent lines with stable resistance to bacterial wilt were selected.

[0017] In the embodiment of the present application, by continuously collecting and time-series segmenting the morphological characteristics, physiological indicators, and disease symptoms during the growth period of peanuts, a complete growth cycle feature library was constructed, realizing the comprehensive monitoring and recording of the plant growth and development process. By comparing and analyzing the phenotypic trajectories of healthy plants and diseased plants, an early warning index system was established, providing data support for timely discovery of potential susceptible individuals. A deep learning algorithm was introduced into the bacterial wilt early warning model, and the time series analysis ability of the recurrent neural network was used to monitor and analyze the growth phenotypic data in real time, improving the accuracy of early warning. By tracking the entire growth period of plants with different resistance levels and comparing the phenotypic characteristics at the critical period, a dynamic resistance evaluation scheme was established, making the evaluation of bacterial wilt resistance more objective and comprehensive. Through the environmental response analysis of multi-point test data and the evaluation of plant-environment interaction, an adaptability index system was constructed, enhancing the stability of line selection. Finally, a comprehensive evaluation of the breeding materials was carried out through a hierarchical screening system, realizing the precise screening of excellent plant lines. This method applies artificial intelligence algorithms to the field of peanut bacterial wilt resistance breeding. Through the long short-term memory network, feature extraction and time series analysis of plant growth data are carried out, fully mining the bacterial wilt resistance information contained in the growth cycle data, providing a data-driven scientific decision-making basis for the breeding of bacterial wilt resistant varieties. At the same time, this method establishes a complete data processing and analysis process, realizing the whole process intelligence from data collection, feature extraction, early warning analysis to line selection, greatly improving the efficiency and accuracy of the bacterial wilt resistance breeding work.

[0018] In a specific embodiment, the process of executing step S101 may specifically include the following steps: (1)Collect the morphological characteristics in a periodic segmented manner. Divide the morphological characteristics into the seedling stage, vegetative growth stage, and reproductive growth stage, and quantitatively measure the plant height, number of branches, and leaf area for each growth stage. (2)Extract the time-series change data of photosynthesis intensity, respiration intensity, and chlorophyll content from the physiological indicators, and perform standardization processing and outlier removal on the data. (3)Rate and score the disease symptoms, and establish a symptom index data set including lesion size, disease index, and incidence rate. (4)Integrate the segmented data of morphological characteristics, time-series change data, and symptom index data to construct a multi-dimensional data set reflecting the characteristics of different growth stages. (5)Extract features from the multi-dimensional data set through data dimensionality reduction processing, and establish a feature vector including key growth stage indicators. (6)Sort and organize the feature vectors in time series, construct a data association network covering the entire growth period, and obtain a growth cycle feature library.

[0019] Specifically, as Figure 3 shown, it is a schematic flowchart of the process of dividing data according to key growth stages through time series segmentation, constructing time series feature vectors, and obtaining a growth cycle feature library in the embodiment of the present application. Collect the morphological feature data, physiological index data, and disease symptom data at a specified time period. After the collected data is subjected to standardization processing and outlier removal, a multi-dimensional data set is formed through time axis alignment, and then feature dimensionality reduction is performed through principal component analysis. Finally, a data association network is constructed in chronological order to form a complete growth cycle feature library.

[0020] Using the mutagenized material as the female parent and the elite variety as the male parent to configure a hybrid combination to obtain F1 generation individuals, the entire growth period is divided into three stages: the seedling stage (15 - 30 days after sowing), the vegetative growth stage (31 - 60 days after sowing), and the reproductive growth stage (61 - 120 days after sowing). During each growth stage, use a vernier caliper to measure the plant height, and use a plant height measuring instrument to collect plant height data for each plant, collecting 2 times a week; use the leaf counting method to count the number of branches and record the number of main stem branches and lateral branches; measure the leaf area index through a leaf area meter, and select the functional leaves of each plant for measurement. Establish an independent data table for the data within each growth stage. The data table includes three dimensions: plant number, measurement time, and measurement value.

[0021] The collection of physiological indicators focuses on the photosynthetic capacity, respiratory metabolism, and changes in chlorophyll content of plants. The photosynthetic intensity is measured using a photosynthesis meter, and the measurement is carried out at 9:00 - 11:00 in the morning, recording the net photosynthetic rate; the respiratory intensity is obtained by measuring the oxygen consumption rate, and the clark oxygen electrode method is used for measurement; the chlorophyll content is measured in situ using a chlorophyll meter. The collected original data is standardized, and the z-score standardization method is used to unify data with different dimensions to the same scale. Outlier detection is performed on the standardized data, and the box plot method is used to identify outliers. Data points exceeding 1.5 times the interquartile range are marked as outliers and removed. The disease symptom rating uses a 0 - 9 level standard, where level 0 indicates no disease spots and level 9 indicates the whole plant is dead. The size of the disease spots is measured using a ruler to measure the length and width of the disease spots, and the disease spot area is calculated; the disease index is calculated based on the degree of disease, and the disease index = (∑(number of diseased plants at each level × corresponding level) / (total number of plants surveyed × highest level)) × 100; the incidence rate is obtained by calculating the ratio of the number of diseased plants to the total number of plants. These indicator data are stored in the symptom indicator dataset, and the dataset contains fields such as time, plant number, symptom level, measurement value, etc.

[0022] In the data integration stage, the segmented data of morphological characteristics, the time-series data of physiological indicators, and the disease symptom data are correlated. A unified data format is established, and each record contains a timestamp, a plant identifier, and the values of each indicator. The data alignment method is used to map the data collected at different time points to a unified time axis to form a complete multi-dimensional dataset. Each row in the multi-dimensional dataset represents an observation point, and the columns contain all the collected characteristic indicators. Feature extraction uses the principal component analysis method to reduce the high-dimensional data to a lower dimension while retaining the key information. Calculate the feature covariance matrix, solve for the eigenvalues and eigenvectors, and select the principal components with larger contribution rates as the key features. The selection of the principal components is based on the principle that the cumulative contribution rate reaches more than 85%. The selected principal components are used to recombine the original data according to the corresponding eigenvectors to obtain the feature vectors after dimensionality reduction.

[0023] Arrange the feature vectors in chronological order to construct a data association network reflecting the entire process of plant growth and development. The nodes in the network structure represent the feature states at different times, and the edges represent the time-series association relationships between the features. By calculating the correlation coefficients between the features, the feature connection weights are established to form a complete growth cycle feature library.

[0024] For example: Select the anti-bacterial wilt female parent variety A and the high-yield male parent variety B for hybridization to obtain the F1 generation population, and start data collection at the seedling stage. The plant height data of each plant at the seedling stage contains 6 measurement points, and the measured plant height data is processed by z-score standardization. The simultaneously collected photosynthesis intensity data shows a diurnal variation pattern, and the trend data is obtained by eliminating short-term fluctuations through the moving average method. The disease symptom rating shows that some plants have lesions during the vegetative growth period, and the lesion size and expansion rate are recorded. Align these data according to time to construct a data matrix containing 15 feature dimensions. Through principal component analysis, the feature dimensions are reduced to 5 principal components, which reflect the main characteristics of plant growth, metabolism, and resistance to bacterial wilt. Organize these features in a time series to form a growth cycle data record.

[0025] In a specific embodiment, the process of executing step S102 may specifically include the following steps: (1) Extract the morphological characteristics, physiological indicators, and disease symptom data of healthy plants from the growth cycle feature library to establish a phenotypic dataset of healthy plants; (2) Extract the morphological characteristics, physiological indicators, and disease symptom data of diseased plants from the growth cycle feature library to establish a phenotypic dataset of diseased plants; (3) Conduct a time series comparison between the phenotypic dataset of healthy plants and the phenotypic dataset of diseased plants, calculate the feature difference values at each time point, and construct a phenotypic trajectory difference matrix; (4) Segment the phenotypic trajectory difference matrix by growth stage, statistically analyze the distribution of differential characteristics in each growth stage, and form a differential feature statistic; (5) Set a difference threshold based on the differential feature statistic, screen out the significant differential characteristics in the critical growth stages, and construct an early warning indicator set; (6) Input the early warning indicator set into the warning evaluation network constructed by a recurrent neural network. The warning evaluation network consists of an input layer, a feature extraction layer, a time series analysis layer, and a warning output layer; The input layer contains N input nodes, where N is the feature dimension of the early warning indicators, and each node corresponds to a phenotypic feature data; the feature extraction layer is composed of long short-term memory units, and each unit contains a forgetting gate, an input gate, and an output gate, which are used to capture the long-term dependence relationship of phenotypic features; the time series analysis layer adopts a bidirectional structure, processing the feature sequence from the current moment to the future moment in the forward direction and the feature sequence from the historical moment to the current moment in the reverse direction to achieve the analysis of the whole process data of the growth stage; the warning output layer calculates the probability distribution of different risk levels through the Softmax function, compares the probability values with the preset threshold to generate a warning decision rule, and obtains an anti-bacterial wilt early warning model.

[0026] Specifically, when classifying and extracting data from the growth cycle feature library, plants are classified according to the disease symptom rating results. Plants with a disease symptom rating of 0-2 are classified as healthy plants, and plants with a disease symptom rating above 3 are classified as diseased plants. For the healthy plant group, morphological characteristics (plant height, number of branches, leaf area), physiological indicators (photosynthesis intensity, respiration intensity, chlorophyll content), and disease symptom (lesion size, disease index, incidence rate) data are extracted to establish a healthy plant phenotype dataset. The above-mentioned index data are extracted from the diseased plant group using the same method to establish a diseased plant phenotype dataset. When making a time-series comparison of the two phenotype datasets, the data are aligned along the time axis to ensure the comparability of the two sets of data at the same time point. Calculate the characteristic difference value between the healthy group and the diseased group at each time point. The difference value is calculated using the relative change amount, that is, (diseased group value - healthy group value) / healthy group value. Arrange the calculated difference values in chronological order to construct a phenotype trajectory difference matrix. The rows of the matrix represent the time series, and the columns represent different characteristic indicators.

[0027] When segmenting the phenotype trajectory difference matrix, the data are segmented according to the three growth stages of the seedling stage, vegetative growth stage, and reproductive growth stage. Statistical analysis is performed on the differential characteristics within each growth stage, and statistics such as the mean, standard deviation, and coefficient of variation of each characteristic indicator within that stage are calculated. In this way, characteristic difference statistics reflecting the distribution law of characteristic differences in different growth stages are obtained.

[0028] When setting the difference threshold based on the feature difference statistic, a statistical significance test method is adopted. A normality test is performed on the feature difference statistic to determine the data distribution characteristics. According to the test results, a parametric test or non-parametric test method is selected to calculate the significance level. The features with a significance level reaching 0.05 are marked as significantly different features. For the significantly different features, their performance characteristics at different growth stages are further analyzed, and the features that show significant differences at an early stage are selected as early warning indicators to construct an early warning indicator set. When the early warning indicator set is input into the warning evaluation network, the input layer of the network is constructed. The number of nodes N in the input layer is equal to the dimension number of the early warning indicators, and each node corresponds to a feature indicator. The feature extraction layer adopts a long short-term memory unit (LSTM) structure, and each LSTM unit contains three gating units: a forget gate, an input gate, and an output gate. The forget gate is responsible for determining the information to be discarded from the cell state, the input gate determines the new information to be updated, and the output gate determines the output cell state. The time series analysis layer adopts a bidirectional structure, including a forward and a backward LSTM layer. The forward LSTM layer processes the sequence information from the current moment to the future moment, and the backward LSTM layer processes the sequence information from the historical moment to the current moment. The warning output layer adopts a Softmax activation function to convert the network output into a probability distribution of different risk levels. The probability value is compared with a preset warning threshold to generate a warning determination result.

[0029] For example, in the F2 generation population of peanut breeding for bacterial wilt resistance, 75 healthy plants and 25 diseased plants are identified from 100 plants. Growth period data are extracted from these two groups of plants respectively, including phenotypic characteristic records at 15 time points. Through time series comparison analysis, it is found that at the initial stage of the vegetative growth period, there is a significant difference in chlorophyll content between the healthy group and the diseased group, and this difference gradually expands in subsequent periods. This feature is incorporated into the early warning indicator set. When constructing the warning evaluation network, 12 nodes are set in the input layer, corresponding to 12 early warning indicators. The number of LSTM units in the feature extraction layer is set to 64, and the time series analysis layer adopts a bidirectional structure with 32 LSTM units. After network training, a new plant sample is evaluated for warning. According to the probability distribution of the Softmax output layer, the warning threshold is set to 0.7. When the disease risk probability of a certain plant exceeds this threshold, it is marked as a potentially susceptible individual.

[0030] In a specific embodiment, the process of executing step S103 may specifically include the following steps: (1) Input the morphological characteristics, physiological indexes, and disease symptom data of the breeding materials into the N nodes of the input layer, and perform standardization processing on the phenotypic characteristic data of each node to obtain standardized growth phenotypic data; (2) Input the standardized growth phenotype data into the long short-term memory unit of the feature extraction layer. Filter historical information through the forget gate, receive current information through the input gate, and integrate information through the output gate to obtain the time-series correlation data of phenotypic features; (3) Input the time-series correlation data of phenotypic features into the time-series analysis layer, and process the feature sequences from the current moment to the future moment and from the historical moment to the current moment through a bidirectional structure to obtain the analysis data for the entire growth period; (4) Calculate the analysis data for the entire growth period through the Softmax function of the early warning output layer to obtain the probability distribution values of different risk levels; (5) Compare the probability distribution values with the early warning threshold, mark the potential susceptible individuals exceeding the threshold, and construct the risk monitoring results; (6) Conduct feature analysis and risk level ranking on the susceptible individuals in the risk monitoring results to obtain the risk early warning results.

[0031] Specifically, the input data processing of the early warning evaluation network extracts morphological features (plant height, number of branches, leaf area), physiological indicators (photosynthesis intensity, respiration intensity, chlorophyll content), and disease symptoms (lesion size, disease index, incidence rate) data from breeding materials. These data are input into N input nodes of the early warning network, and the value of N is equal to the total number of feature dimensions. Standardize the input data using the z-score standardization method to convert each indicator to a unified scale.

[0032] The standardized growth phenotype data is input into the long short-term memory unit of the feature extraction layer. Each long short-term memory unit contains three gating units: the forget gate, the input gate, and the output gate. The forget gate selectively forgets the cell state at the previous moment, mainly filtering out historical information with low contribution to the current early warning judgment; the input gate receives new information at the current moment, focusing on the change amount of phenotypic features; the output gate integrates the filtered historical information with the current information to generate the time-series correlation data of phenotypic features. The time-series correlation data reflects the change law of the plant's phenotypic features over time.

[0033] Input the time-series correlation data of phenotypic features into the time-series analysis layer, and this layer uses a bidirectional structure for data processing. The forward processing process analyzes the feature change trend from the current moment to the predicted future moment, and the backward processing process analyzes the feature evolution law from the historical moment to the current moment. Obtain the complete analysis data for the entire growth period through bidirectional analysis, and these data contain the growth and development characteristics of the plant during the entire growth cycle and the information on the resistance to bacterial wilt.

[0034] Calculate the risk level probability for the analysis data of the entire growth period through the Softmax function. The calculation formula is: Wherein: represents the probability value of the i-th risk level; represents the number of characteristics throughout the growth period; represents the temporal weight of the j-th characteristic; represents the contribution degree of the j-th characteristic to bacterial wilt resistance; represents the growth cycle coefficient of the j-th characteristic; represents the total number of risk levels.

[0035] Compare the calculated probability distribution value with the preset warning threshold. The warning threshold is a critical value determined based on historical data statistics. When the probability value of a certain risk level exceeds the corresponding threshold, mark the plant as a potentially susceptible individual. Establish a risk monitoring result dataset for all marked susceptible individuals. Further analyze the susceptible individuals in the risk monitoring results, count the characteristic performances of the susceptible individuals, calculate the deviation degree of each characteristic index from healthy plants, and rank the susceptible individuals according to the deviation degree to form a risk warning result.

[0036] For example, in the selection process of the F2 generation population in peanut breeding for bacterial wilt resistance, input the phenotypic data of the hybrid offspring of mutagenic materials with strong comprehensive bacterial wilt resistance and excellent varieties into the warning evaluation network. Standardize the data of 9 characteristic indexes such as plant height and photosynthesis intensity. The standardized data is input into the LSTM unit of the feature extraction layer, and the temporal features of the data are processed through the gating mechanism. For example, for the chlorophyll content index, the forgetting gate filters out the short-term changes of day-night fluctuations and retains the long-term change trend; the input gate focuses on the mutation points of the chlorophyll content; the output gate integrates this information to form the temporal features reflecting the change law of the chlorophyll content. The bidirectional structure of the temporal analysis layer processes the historical performance of the plant from emergence to the current moment and predicts the future growth and development trend respectively. Based on the processed data, calculate the probability that the plant belongs to different risk levels through the Softmax function. Compare the risk probability with the warning threshold, identify the potential susceptible individuals, and determine the risk level according to the abnormal degree of each characteristic index, and screen out the individuals suitable for continuing to participate in breeding.

[0037] In a specific embodiment, the process of executing step S104 may specifically include the following steps: (1) Divide the resistance levels of plants according to the risk warning results, divide the plants into four levels: high resistance, medium resistance, slight susceptibility, and high susceptibility, and establish a hierarchical dataset; (2) Conduct full-growth-period tracking and monitoring on plants with different resistance levels in the hierarchical dataset, collect the morphological characteristics, physiological indexes, and disease symptom data at each key growth stage, and construct a growth-period tracking data matrix; (3) Extract the characteristic data of each critical period from the growth period tracking data matrix, compare the characteristics of the plants within the same resistance level group, and obtain the within-group characteristic consistency index; (4) Compare and analyze the within-group characteristic consistency index among different resistance level groups, calculate the significance of characteristic differences, and screen out the phenotypic characteristics of critical periods; (5) Establish a scoring standard for the phenotypic characteristics of critical periods, divide the characteristic values into different scores according to numerical intervals, and form a resistance scoring rule; (6) Score and calculate the phenotypic characteristics of the plants based on the resistance scoring rule, and synthesize the characteristic scores of each item to obtain a dynamic resistance evaluation scheme.

[0038] Specifically, when classifying the resistance levels of plants according to the risk warning results, a four-level classification standard is adopted. The high resistance level corresponds to plants with a risk warning value less than 0.3, showing no obvious disease symptoms throughout the growth period; the medium resistance level corresponds to plants with a risk warning value between 0.3 and 0.5, showing only mild disease symptoms in the later growth stage; the low susceptibility level corresponds to plants with a risk warning value between 0.5 and 0.7, showing disease symptoms starting from the middle growth stage; the high susceptibility level corresponds to plants with a risk warning value greater than 0.7, showing obvious disease symptoms in the early growth stage. Separate datasets are established for the plants in the four levels, recording the plant numbers, resistance levels, and risk warning values.

[0039] The whole growth period of the classified plant population is tracked and monitored, and key monitoring points are set at the seedling stage, vegetative growth stage, and reproductive growth stage respectively. At each monitoring point, data on morphological characteristics (plant height, number of branches, leaf area), physiological indicators (photosynthesis intensity, respiration intensity, chlorophyll content), and disease symptoms (lesion size, disease index, incidence rate) are collected. The collected data are sorted into a growth period tracking data matrix according to time sequence and characteristic type. The rows of the matrix represent different time points, and the columns represent different characteristic indicators. Extract the characteristic data of each critical growth period from the growth period tracking data matrix. For the plants within the same resistance level group, calculate the mean and standard deviation of each characteristic indicator to evaluate the characteristic consistency of the plants within the group. The coefficient of variation is used for consistency evaluation. The smaller the coefficient of variation, the more consistent the performance of the plants within the group. For each characteristic indicator, calculate the significance of characteristic differences among the four resistance level groups, use the one-way analysis of variance method to calculate the F statistic and P value, and the characteristics with a P value less than 0.05 are marked as significantly different characteristics.

[0040] For the establishment of the scoring standard for the phenotypic characteristics of critical periods, the following formula is adopted: Where: represents the score of the k-th plant at the j-th resistance level in the i-th growth period; Represents the number of characteristic indicators; Represents the weight coefficient of the m-th characteristic; Represents the actual measured value; Represents the minimum value of this characteristic; Represents the maximum value of this characteristic; Represents the effectiveness coefficient of the characteristic.

[0041] Formula for calculating the comprehensive resistance score based on the scoring rules: Among them: Represents the comprehensive resistance score of the k-th plant; Represents the number of growth stages; Represents the number of resistance levels; Represents the temporal weight of the i-th growth stage; Represents the level coefficient of the j-th resistance level; Represents the interaction coefficient of growth stage - resistance level.

[0042] Organize the calculated comprehensive scoring results into a dynamic resistance evaluation plan. For example: In the selection of the F2 generation population, select individuals with different performances from the plants planted in the bacterial wilt spectrum for grading. When tracking the growth stages of the highly resistant group of plants, it is recorded that the chlorophyll content of these plants remains at a relatively high level during the vegetative growth period, and the coefficient of variation is small, indicating that this characteristic has good consistency within the highly resistant group. Further comparing the chlorophyll content data of different resistance level groups, it is found that there are significant differences in this indicator among different resistance levels. Incorporate the chlorophyll content indicator into the scoring standard and set the corresponding weight coefficient. Through the comprehensive scoring of multiple periods and multiple characteristic indicators, a batch of individuals with stable performance against bacterial wilt are selected, and these individuals enter the F3 generation southern breeding and generation-adding link to continue the resistance verification.

[0043] In a specific embodiment, the process of executing step S105 may specifically include the following steps: (1) Collect the plant growth data in the dynamic resistance evaluation plan at different geographical locations, record the growth conditions, soil characteristics, and climate parameters, and construct a multi-point test data set; (2) Conduct site-by-site statistics on the multi-point test data set, calculate the correlation degree between the plant phenotypic data and environmental factors at each test site, and form an environmental response data matrix; (3) Extract the temporal variation law of plant growth and environmental factors from the environmental response data matrix, and establish a plant - environment interaction data table; (4) Perform variance decomposition on the data in the plant - environment interaction data table, separate the genotype effect, environmental effect, and interaction effect, and obtain the effect component data; (5) Calculate the performance stability of each genotype based on the effect component data, and construct an adaptability index system that includes genotype-by-environment interaction; (6) Conduct numerical analysis on the adaptability index system, screen out genotype combinations with high stability, and obtain the strain stability parameters.

[0044] Specifically, during the regional trial stage, experimental sites are set in representative ecological regions, including cold regions, temperate regions, and subtropical regions. Detailed environmental parameters are recorded at each experimental site, including growth conditions (light intensity, sunshine duration, precipitation), soil properties (pH value, organic matter content, trace elements), and climate parameters (temperature, humidity, wind speed). At the same time, phenotypic data of the planted peanut strains are collected, recording growth and development indicators and the performance of resistance to bacterial wilt, and these data are organized into a multi-site experimental data set. When analyzing the multi-site experimental data set, the data is grouped and processed according to the experimental site positions. The pearson correlation coefficient is used to calculate the correlation between the plant phenotypic data and environmental factors, and a correlation matrix is established. Each element in the correlation matrix represents the correlation intensity between a phenotypic index and an environmental factor, and the value range of the correlation coefficient is from -1 to 1, with a positive value indicating a positive correlation and a negative value indicating a negative correlation.

[0045] When extracting the temporal variation pattern from the environmental response data matrix, the data is divided according to the growth stages. For the data within each stage, the dynamic correlation between the plant phenotype and environmental factors is calculated, and the degree of influence of environmental factor changes on plant growth and development is analyzed. The analysis results are organized into a plant-environment interaction data table, which contains the interaction relationships at different growth stages.

[0046] Perform variance decomposition on the plant-environment interaction data table using the two-way analysis of variance method. Decompose the total variance into three parts: genotype effect (differences between different strains), environmental effect (differences between different experimental sites), and interaction effect (differences in the response of strains to the environment). By calculating the variance components of each part, the degree of influence of different factors on plant performance is evaluated.

[0047] The formula for calculating the stability index based on the effect component data is: Where: represents the stability index of the i-th genotype for the k-th trait in the j-th environment; represents the number of environmental points; represents the number of test periods; represents the environmental weight coefficient; represents the period weight coefficient; represents the measured value; represents the mean value; represents the standard deviation.

[0048] The formula for calculating the comprehensive stability parameter is as follows: Where: represents the comprehensive stability parameter; represents the number of genotypes; represents the number of traits; represents the genotype weight; represents the trait weight; represents the coefficient of variation.

[0049] For example: The excellent lines obtained from the F3 generation's southern multiplication and generation addition enter the multi-point test stage. The same batch of materials is planted at test sites in three different climate regions, and during the test period, the temperature change curve, precipitation distribution, and soil nutrient status of each test site are recorded. Through phenotypic data collection, it is found that there are differences in morphological characteristics such as plant height and number of branches of a certain line at different test sites. Through environmental correlation analysis, it is determined that these differences are significantly correlated with environmental factors such as average daily temperature and precipitation. Further through variance decomposition, it is found that the phenotypic variation of this line mainly comes from environmental effects, and the genotype effects and interaction effects are relatively small. According to the stability index, excellent lines with stable performance are selected, and these lines show strong environmental adaptability and stability against bacterial wilt.

[0050] In a specific embodiment, the process of executing step S106 may specifically include the following steps: (1) Conduct numerical analysis on the line stability parameter, establish an evaluation matrix including stability index, bacterial wilt resistance index, and yield index, and construct a comprehensive evaluation system for breeding materials; (2) Quantitatively calculate each index in the comprehensive evaluation system for breeding materials according to the weight to form a multi-dimensional evaluation score, and establish a comprehensive scoring data set; (3) Conduct clustering processing on the comprehensive scoring data set, divide the breeding materials into different grades according to the score distribution, and form a hierarchical evaluation result; (4) Set a screening threshold for the hierarchical evaluation result, include the materials higher than the threshold in the candidate set of excellent lines, and construct a database of excellent materials; (5) Analyze the growth cycle characteristics of the lines in the database of excellent materials, evaluate the bacterial wilt resistance performance and growth status during the whole growth period, and form a line characteristic portrait; (6) Screen excellent lines based on the line characteristic portrait, comprehensively consider the growth cycle performance and bacterial wilt resistance, and obtain a bacterial wilt resistance breeding plan based on the growth cycle.

[0051] Specifically, the numerical analysis of the strain stability parameters requires the establishment of a complete evaluation index system. The stability index includes the genotype effect value, the environmental effect value, and the interaction effect value, which reflect the stability degree of the plant's performance under different environments; the bacterial wilt resistance index includes the disease incidence rate, the disease index, the lesion size, and the disease progression rate, which reflect the plant's ability to resist bacterial wilt; the yield index includes the yield per plant, the number of pods, the seed setting rate, and the weight of 100 fruits, which reflect the plant's yield performance. These indexes are organized into an evaluation matrix, where the rows of the matrix represent different breeding materials and the columns represent different evaluation indexes. In the quantification calculation stage, different weights are assigned to each index in the evaluation matrix. The weight of the stability index is set to 0.4. Considering the core goal of bacterial wilt resistance breeding, the weight of the bacterial wilt resistance index is set to 0.4, and the weight of the yield index is set to 0.2. Corresponding weight coefficients are also set for the specific indexes within each index category. After standardizing the original data, it is multiplied by the weight coefficient to obtain the weighted score. The weighted scores of all indexes are summarized to obtain the comprehensive evaluation score of each breeding material, and a comprehensive scoring data set is established.

[0052] When performing cluster analysis on the comprehensive scoring data set, the K-means clustering algorithm is used to divide the breeding materials into four grades: excellent, good, average, and poor. During clustering, the comprehensive score is used as the feature dimension, the Euclidean distance between samples is calculated, and the clustering center is determined through iterative optimization to obtain the hierarchical evaluation result.

[0053] Screening thresholds are set for the hierarchical evaluation results, and the materials in the excellent grade are selected as candidate excellent strains. Materials with a comprehensive score higher than 90 points are directly included in the candidate set of excellent strains. Materials with scores between 80 and 90 points need to be further evaluated to see if their key indexes meet the requirements. The information of the selected materials is stored in the excellent material database, which records information such as the material number, the comprehensive score, and the values of each index. The whole growth period analysis is carried out on the strains in the excellent material database, with a focus on the performance in three key stages: the seedling stage, the vegetative growth stage, and the reproductive growth stage. Record the morphological characteristics (plant height, number of branches, leaf area), physiological indexes (photosynthesis intensity, respiration intensity, chlorophyll content), and bacterial wilt resistance performance (lesion size, disease index, incidence rate) at each stage. These data are organized into a time series to construct a characteristic portrait reflecting the whole process of plant growth and development.

[0054] When screening based on the strain characteristic portrait, three aspects are mainly considered: growth vigor, bacterial wilt resistance, and stability. Growth vigor mainly depends on whether the morphological development of the plant is robust; bacterial wilt resistance focuses on the occurrence and development of the disease; stability is to evaluate the consistency of the performance of various traits. By comprehensively analyzing these characteristics, excellent strains are screened out to formulate a bacterial wilt resistance breeding plan.

[0055] For example: After obtaining the strain stability parameters from multi-location trials, a comprehensive evaluation was conducted on a batch of F3 generation materials. An evaluation matrix was established, including 15 evaluation indicators. The weight coefficients of each indicator were determined through preliminary data analysis. For example, the weight of the disease index in the indicator of resistance to bacterial wilt was set to 0.4 because this indicator has the highest correlation with resistance to bacterial wilt. After standardizing and weighting the original data, the comprehensive score of each material was obtained. The K-means clustering method was used to divide the materials into four grades, and a screening threshold of 90 for the comprehensive score was set for excellent strain selection. The growth period of the selected excellent strains was tracked, and the key growth indicators were recorded. Based on the growth characteristic portraits of the materials, excellent strains with both stable resistance to bacterial wilt and good agronomic traits were selected, and these strains entered the regional trial stage.

[0056] In a specific embodiment, the process of performing the step of extracting stability index data in the evaluation system, including the genotype effect value in the strain stability parameters, may specifically include the following steps: (1) Extract stability index data from the comprehensive evaluation system of breeding materials, including the genotype effect value, environmental effect value, and interaction effect value in the strain stability parameters, and construct a stability score matrix; (2) Extract data on the indicator of resistance to bacterial wilt from the comprehensive evaluation system of breeding materials, including the disease incidence rate, disease index, lesion size, and disease progress rate, and construct a score matrix for resistance to bacterial wilt; (3) Extract data on the yield indicator from the comprehensive evaluation system of breeding materials, including the yield per plant, number of pods, seed setting rate, and weight of 100 fruits, and construct a yield score matrix; (4) Set weight coefficients for the stability score matrix, the score matrix for resistance to bacterial wilt, and the yield score matrix, and construct a weight allocation plan; (5) Perform weighted calculations on the score matrices based on the weight allocation plan, integrate the scores of each indicator, and form a comprehensive score table; (6) Perform normalization processing on the comprehensive score table to unify each indicator to the same scale and obtain a comprehensive score data set.

[0057] Specifically, the comprehensive evaluation of breeding materials starts with the extraction of stability index data. Genotype effect values, environmental effect values, and interaction effect values are extracted from the strain stability parameters, which are obtained through variance decomposition. The genotype effect value reflects the contribution degree of the plant's own genetic characteristics, the environmental effect value represents the influence degree of external environmental conditions, and the interaction effect value reflects the plant's response ability to the environment. These three effect values are sorted into a stability scoring matrix according to the material numbers. The rows of the matrix represent different breeding materials, and the columns correspond to the three effect values respectively. The extraction of bacterial wilt resistance index data mainly focuses on four key indicators: the disease incidence rate is obtained by counting the ratio of the number of diseased plants to the total number of plants, the disease index is calculated according to the disease severity level, the lesion size is measured and recorded with a ruler, and the disease progress rate is determined by continuously observing the lesion expansion speed. These indicators need to be recorded during the seedling stage, vegetative growth stage, and reproductive growth stage, and the dynamic process of disease development is reflected through the comparison of data at different times. These indicator data are sorted into a bacterial wilt resistance scoring matrix, with each row representing a material and the four columns corresponding to the four bacterial wilt resistance indicators respectively.

[0058] The extraction of yield index data mainly includes: the yield per plant is obtained by weighing, the number of pods is counted manually, the seed setting rate is calculated according to the ratio of the number of filled pods to the total number of pods, and the weight of 100 fruits is obtained by weighing 100 seeds. These indicator data are mainly collected at the harvest stage and sorted into a yield scoring matrix according to the material numbers. The matrix structure is similar to the previous two matrices, with rows representing materials and columns corresponding to the four yield indicators. When setting the weight coefficients for the three scoring matrices, the balance between the goal of bacterial wilt resistance breeding and production requirements is considered. The weight of the stability scoring matrix is set to 0.3 because stability is the basis for variety promotion; the weight of the bacterial wilt resistance scoring matrix is set to 0.4 as the core indicator for bacterial wilt resistance breeding; the weight of the yield scoring matrix is 0.3 to ensure that the selected breeding materials have good yield performance. Within each matrix, corresponding weights are also set for different indicators. For example, in the bacterial wilt resistance indicators, the disease index has a higher weight because it comprehensively reflects the severity of the disease.

[0059] When performing weighted calculations based on the weight distribution scheme, the original data in each matrix is multiplied by the corresponding indicator weights to obtain the weighted indicator scores. Then, the weighted scores of the same material in different matrices are added together to obtain the comprehensive score of the material. The scores of all materials are sorted into a comprehensive score table, which includes the material numbers and the corresponding comprehensive scores.

[0060] When normalizing the comprehensive score table, the maximum-minimum normalization method is used. The scores of each indicator are converted into the range of 0 - 1 to eliminate the influence of the dimension between different indicators. Through normalization, a standardized comprehensive scoring data set is obtained, which is convenient for subsequent material grading and screening.

[0061] For example, during the selection process of the F3 generation, a comprehensive evaluation was conducted on a batch of materials that had undergone multi-point trials. The effect values were extracted from the strain stability parameters, and it was found that the genotype effect value of a certain material was dominant, indicating that its trait performance was mainly determined by genetic factors. In the evaluation of resistance to bacterial wilt, the disease data of this material was continuously tracked, and it was recorded that the lesion expansion rate was slow and the disease index remained at a low level. At the harvest stage, its yield indicators were measured, and the yield per plant and seed setting rate showed good performance. These data were respectively input into three scoring matrices and calculated according to the set weight coefficients to obtain the comprehensive score of this material. Through this evaluation method, excellent materials with both stability and good resistance to bacterial wilt and yield performance were screened out.

[0062] In a specific embodiment, the process of performing the step of extracting the data of the bacterial wilt resistance index from the comprehensive evaluation system of breeding materials may specifically include the following steps: (1) Divide the disease incidence data into the seedling stage incidence, vegetative growth stage incidence, and reproductive growth stage incidence according to different growth stages, and construct an incidence time series matrix; (2) Classify and count the disease index data according to the lesion area, withering degree, and affected part, and establish a disease grading data table; (3) Extract the time series changes of the lesion length, width, and area from the lesion size data to form a lesion development data set; (4) Conduct time series tracking on the disease progression rate data, calculate the lesion expansion rate and the degree of disease aggravation per unit time, and obtain the disease dynamic parameters; (5) Integrate the incidence time series matrix, the disease grading data table, the lesion development data set, and the disease dynamic parameters to form a bacterial wilt resistance characteristic database; (6) Perform weighted processing on the various indicators in the bacterial wilt resistance characteristic database, calculate the comprehensive score through the combination of weight coefficients, and construct a bacterial wilt resistance scoring matrix.

[0063] Specifically, for the processing of disease data, the disease incidence data is divided according to the growth stage. The statistical time for the seedling stage incidence is 15 - 30 days after sowing, which is obtained by investigating the ratio of the number of diseased plants to the total number of plants; the statistical time for the vegetative growth stage incidence is 31 - 60 days after sowing, with a focus on the increase in newly diseased plants; the statistical time for the reproductive growth stage incidence is 61 - 120 days after sowing, recording the spread of the disease in the population. The incidence data for these three periods is organized into an incidence time series matrix, where the rows in the matrix represent different time points and the columns represent the incidence data for different growth stages. For the processing of the disease index data, a classification and statistics method is used, and ratings are carried out according to three dimensions: lesion area, wilting degree, and damaged part. The lesion area is divided into five levels: level 1 indicates that the lesion area is less than 10% of the leaf area, level 2 is 10 - 25%, level 3 is 26 - 50%, level 4 is 51 - 75%, and level 5 is above 76%; the wilting degree is also divided into five levels: level 1 indicates slight wilting, level 2 is obvious wilting, level 3 is partial wilting, level 4 is mostly wilting, and level 5 is complete withering; the damaged part is divided into three categories: root, stem, and leaf. These grading data are organized into a disease grading data table, which records the disease levels of each plant at different times.

[0064] The processing of lesion size data includes three dimensions: the dynamic changes in lesion length, width, and area. A ruler is used to measure the length and width of the lesion, and the area is calculated according to the lesion shape. Measurements are taken every 3 days to record the spread of the lesion. These measurement data are organized in chronological order to form a dataset reflecting the development process of the lesion. The dataset contains a complete record of the change in lesion size over time. For the processing of the disease progression rate data, two aspects are focused on: the lesion expansion speed and the degree of disease aggravation. The lesion expansion speed is obtained by calculating the increase in lesion area per unit time, and the degree of disease aggravation is represented by the change rate of the disease level. These data are continuously tracked, and the change amount within each time interval is recorded to obtain parameters reflecting the dynamic development of the disease.

[0065] In the data integration stage, correlation analysis is performed on the incidence time series matrix, disease severity data table, lesion development dataset, and disease dynamic parameters. A unified data format is established to align data from different sources to the same time axis. Through correlation analysis, the mutual relationships between different indicators are determined, and a complete database of anti-bacterial wilt resistance characteristics is constructed. When conducting a comprehensive score based on the anti-bacterial wilt resistance characteristics database, weight coefficients are set for each indicator. The weight of the incidence rate indicator is 0.3 because it reflects the anti-bacterial wilt resistance at the population level; the weight of the disease index is 0.3, which reflects the severity of the disease in a single plant; the weight of the lesion development indicator is 0.2, indicating the expansion trend of the disease; the weight of the disease progression rate is 0.2, reflecting the dynamic characteristics of anti-bacterial wilt resistance. The comprehensive anti-bacterial wilt resistance score of each plant is calculated through the combination of these weight coefficients, and an anti-bacterial wilt resistance scoring matrix is constructed.

[0066] For example: Conduct an evaluation of anti-bacterial wilt resistance in the F2 population, and conduct continuous observations starting from sowing. A small number of plants show disease symptoms at the seedling stage, and the initial incidence rate is recorded; after entering the vegetative growth stage, new diseased plants are counted, and it is found that the disease begins to spread in the population; by the reproductive growth stage, the development of the disease is continuously tracked. The disease severity of the diseased plants is classified, and the changes in the lesion size are recorded. The lesion of a certain plant initially appears on the leaves. Through continuous measurement, it is found that the expansion rate of the lesion is slow and the change in the disease severity level is not obvious, indicating that the plant has a certain degree of anti-bacterial wilt resistance. These observed data are input into the scoring system, and the comprehensive anti-bacterial wilt resistance score of the plant is calculated. This score serves as an important reference index in the process of anti-bacterial wilt resistance breeding.

[0067] The above describes the method for analyzing the peanut anti-bacterial wilt resistance breeding strategy based on the growth cycle in the embodiments of the present application. Next, the peanut anti-bacterial wilt resistance breeding strategy analysis system based on the growth cycle in the embodiments of the present application will be described. Please refer to Figure 4 , an embodiment of the peanut anti-bacterial wilt resistance breeding strategy analysis system based on the growth cycle in the embodiments of the present application includes: A collection module 201, configured to continuously collect the morphological characteristics, physiological indicators, and disease symptoms of peanuts during the growth period through a phenotypic monitoring system, divide the data according to key growth stages through time series segmentation, construct a time series feature vector, and obtain a growth cycle feature library; A comparison module 202, configured to perform a comparative analysis on the phenotypic trajectories of healthy plants and diseased plants according to the growth cycle feature library, establish an early warning indicator through differential feature extraction, and obtain an anti-bacterial wilt resistance early warning model; A monitoring module 203, configured to, based on the anti-bacterial wilt resistance early warning model, perform real-time monitoring and analysis on the growth phenotypic data of breeding materials, determine potential susceptible individuals through the warning threshold determination, and obtain a risk warning result; A tracking module 204, configured to perform full-course tracking on plants with different resistance levels during the growth period based on the risk warning result, establish a resistance scoring system through comparison of phenotypic characteristics at key stages, and obtain a dynamic resistance evaluation scheme; An analysis module 205, configured to collect multi-point test data for environmental response analysis according to the dynamic resistance evaluation scheme, establish adaptability indicators through plant-environment interaction evaluation, and obtain line stability parameters; A grading module 206, configured to comprehensively evaluate and grade breeding materials according to the line stability parameters, determine excellent plant lines through a grading and screening system, and obtain a breeding scheme for resisting bacterial wilt based on the growth cycle.

[0068] Through the collaborative cooperation of the above-mentioned various components, by continuously collecting and performing time series segmentation on the morphological characteristics, physiological indicators, and disease symptoms of peanuts during the growth period, a complete growth cycle feature library is constructed, realizing comprehensive monitoring and recording of the plant growth and development process. Through comparative analysis of the phenotypic trajectories of healthy plants and diseased plants, an early warning index system is established, providing data support for timely discovery of potential susceptible individuals. A deep learning algorithm is introduced into the bacterial wilt early warning model, and the time series analysis ability of the recurrent neural network is used to monitor and analyze the growth phenotype data in real time, improving the accuracy of early warning. Through full-course tracking of the growth period of plants with different resistance levels and comparison of phenotypic characteristics at key stages, a dynamic resistance evaluation scheme is established, making the evaluation of bacterial wilt resistance more objective and comprehensive. Through environmental response analysis of multi-point test data and plant-environment interaction evaluation, an adaptability index system is constructed, enhancing the stability of line selection. Finally, through comprehensive evaluation of breeding materials by a grading and screening system, accurate screening of excellent plant lines is realized. This method applies artificial intelligence algorithms to the field of peanut bacterial wilt resistance breeding. Through long short-term memory networks, feature extraction and time series analysis are performed on plant growth data, fully mining the information on bacterial wilt resistance contained in the growth cycle data, providing a data-driven scientific decision-making basis for the breeding of bacterial wilt-resistant varieties. At the same time, this method establishes a complete data processing and analysis process, realizing the full-process intelligence from data collection, feature extraction, early warning analysis to line screening, greatly improving the efficiency and accuracy of the bacterial wilt resistance breeding work.

[0069] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described system, system, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0070] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than limiting them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A method for analyzing peanut bacterial wilt resistance breeding strategies based on growth cycles, characterized in that: The peanut bacterial wilt resistance breeding strategy analysis method based on the growth cycle includes: The morphological characteristics, physiological indicators and disease symptoms of peanuts during their growth period are continuously collected through the phenotypic monitoring system. The data are divided into key growth periods through time series segmentation, and a time series feature vector is constructed to obtain a growth cycle feature library. According to the growth cycle feature library, the phenotypic trajectories of healthy plants and diseased plants are compared and analyzed, and early warning indicators are established by extracting differential features to obtain an early warning model for resistance to bacterial wilt; Based on the bacterial wilt resistance early warning model, the growth phenotype data of the breeding materials are monitored and analyzed in real time, and the potential susceptible individuals are determined by early warning thresholds to obtain risk early warning results; Based on the risk warning results, plants with different resistance levels are tracked throughout their growth period, and a resistance scoring system is established by comparing phenotypic characteristics during key periods to obtain a dynamic resistance evaluation scheme; According to the dynamic resistance evaluation scheme, multi-point test data are collected for environmental response analysis, and adaptability indicators are established through plant-environment interaction evaluation to obtain strain stability parameters; According to the strain stability parameters, the breeding materials are comprehensively evaluated and graded, and excellent strains are determined through a graded screening system to obtain a breeding program for bacterial wilt resistance based on the growth cycle.

2. The method for analyzing peanut bacterial wilt resistance breeding strategies based on growth cycle according to claim 1, characterized in that: The morphological characteristics, physiological indicators and disease symptoms of peanuts during the growth period are continuously collected through the phenotypic monitoring system, and the data are divided according to the key growth period through time series segmentation, and a time series feature vector is constructed to obtain a growth cycle feature library, including: The morphological characteristics are collected periodically and segmented, and the morphological characteristics are divided into seedling stage, vegetative growth stage and reproductive growth stage, and the plant height, branch number and leaf area of ​​each growth stage are quantitatively measured; Extract the time series variation data of photosynthesis intensity, respiration intensity and chlorophyll content from the physiological indicators, standardize the data and remove outliers; Rating and scoring the disease symptoms to establish a symptom indicator data set including lesion size, disease index, and incidence rate; The segmented data of morphological characteristics, temporal variation data and symptom index data are integrated to construct a multidimensional data set reflecting the characteristics of different growth periods; Extracting features from the multidimensional data set through data dimensionality reduction processing to establish a feature vector containing key growth period indicators; The characteristic vectors are sorted and organized in time series, a data association network covering the entire growth period is constructed, and the growth cycle characteristic library is obtained.

3. The method for analyzing peanut bacterial wilt resistance breeding strategies based on growth cycle according to claim 1, characterized in that: According to the growth cycle feature library, the phenotypic trajectories of healthy plants and diseased plants are compared and analyzed, and early warning indicators are established by extracting difference features to obtain an early warning model for resistance to bacterial wilt, including: Extracting morphological characteristics, physiological indicators and disease symptom data of healthy plants from the growth cycle feature library to establish a healthy plant phenotype data set; Extracting the morphological characteristics, physiological indicators and disease symptom data of diseased plants from the growth cycle feature library to establish a phenotypic data set of diseased plants; Performing a time series comparison on the healthy plant phenotypic dataset and the diseased plant phenotypic dataset, calculating the characteristic difference value at each time point, and constructing a phenotypic trajectory difference matrix; The phenotypic trajectory difference matrix is ​​segmented according to the growth period, and the difference characteristic distribution of each growth period is counted to form characteristic difference statistics; Setting a difference threshold based on the characteristic difference statistics, screening significant difference characteristics of key growth periods, and constructing an early warning indicator set; Inputting the early warning indicator set into a warning evaluation network constructed by a recursive neural network, wherein the early warning evaluation network is composed of an input layer, a feature extraction layer, a time series analysis layer, and a warning output layer; The input layer includes N input nodes, where N is the characteristic dimension of the early warning indicator, and each node corresponds to a phenotypic characteristic data; the feature extraction layer is composed of long-term and short-term memory units, each unit includes a forget gate, an input gate and an output gate, which are used to capture the long-term dependency of phenotypic characteristics; the time series analysis layer adopts a bidirectional structure, forwardly processes the feature sequence from the current moment to the future moment, and reversely processes the feature sequence from the historical moment to the current moment, so as to realize the analysis of the data of the entire growth period; the warning output layer calculates the probability distribution of different risk levels through the Softmax function, compares the probability value with the preset threshold value to generate a warning judgment rule, and obtains an anti-bacterial wilt disease warning model.

4. The method for analyzing peanut bacterial wilt resistance breeding strategies based on growth cycle according to claim 3, characterized in that: According to the bacterial wilt resistance early warning model, the growth phenotype data of the breeding materials are monitored and analyzed in real time, and the potential susceptible individuals are determined by early warning thresholds to obtain risk early warning results, including: Inputting the morphological characteristics, physiological indicators and disease symptom data of the breeding materials into the N nodes of the input layer, and performing standardization processing on the phenotypic characteristic data of each node to obtain standardized growth phenotypic data; The standardized growth phenotype data is transmitted to the long short-term memory unit of the feature extraction layer, the historical information is filtered through the forget gate, the input gate receives the current information, and the output gate integrates the information to obtain the phenotypic feature time series correlation data; The phenotypic characteristic time series association data is input into the time series analysis layer, and the characteristic sequences from the current moment to the future moment and from the historical moment to the current moment are processed through a bidirectional structure to obtain the full-process analysis data of the growth period; The whole analysis data of the growth period is calculated through the Softmax function of the early warning output layer to obtain probability distribution values ​​of different risk levels; Comparing the probability distribution value with the warning threshold, marking the potentially susceptible individuals exceeding the threshold, and constructing a risk monitoring result; The susceptible individuals in the risk monitoring results are subjected to feature analysis and risk level ranking to obtain risk warning results.

5. The method for analyzing peanut bacterial wilt resistance breeding strategies based on growth cycle according to claim 1, characterized in that: Based on the risk warning results, plants with different resistance levels are tracked throughout the growth period, and a resistance scoring system is established by comparing phenotypic characteristics during key periods to obtain a dynamic resistance evaluation scheme, including: According to the risk warning results, the resistance levels of the plants are divided into four levels: high resistance, medium resistance, mild sensitivity, and high sensitivity, and a graded data set is established; Tracking and monitoring the plants with different resistance levels in the graded data set throughout their growth period, collecting data on morphological characteristics, physiological indicators and disease symptoms at each key growth period, and constructing a growth period tracking data matrix; Extracting characteristic data of each key period from the growth period tracking data matrix, performing characteristic comparison on plants in the same resistance level group, and obtaining a characteristic consistency index within the group; Comparative analysis is performed on the characteristic consistency index within the group and between groups with different resistance levels, the significance of characteristic differences is calculated, and the phenotypic characteristics of the critical period are screened out; Establishing a scoring standard for the phenotypic characteristics of the critical period, dividing the characteristic values ​​into different scores according to the numerical intervals, and forming a resistance scoring rule; The phenotypic characteristics of the plants are scored and calculated based on the resistance scoring rules, and the scores of various characteristics are combined to obtain a dynamic resistance evaluation scheme.

6. The method for analyzing peanut bacterial wilt resistance breeding strategies based on growth cycle according to claim 1, characterized in that: According to the dynamic resistance evaluation scheme, multi-point test data are collected for environmental response analysis, and adaptability indicators are established through plant-environment interaction evaluation to obtain strain stability parameters, including: Collecting plant growth data in the dynamic resistance evaluation scheme at different geographical locations, recording growth conditions, soil properties, and climate parameters, and constructing a multi-site test data set; Performing point-by-point statistics on the multi-point test data set, calculating the correlation between the plant phenotypic data and the environmental factors at each test point, and forming an environmental response data matrix; Extracting the temporal variation rules of plant growth and environmental factors from the environmental response data matrix, and establishing a plant-environment interaction data table; Performing variance decomposition on the data in the plant-environment interaction data table to separate genotype effects, environmental effects and interaction effects to obtain effect component data; Calculating the performance stability of each genotype based on the effect component data, and constructing an adaptability index system including the interaction between genotype and environment; The adaptability index system is numerically analyzed to screen out genotype combinations with high stability and obtain strain stability parameters.

7. The method for analyzing peanut bacterial wilt resistance breeding strategies based on growth cycle according to claim 1, characterized in that: The method of comprehensively evaluating and grading the breeding materials based on the strain stability parameters, determining the superior strains through a grading screening system, and obtaining a bacterial wilt resistance breeding scheme based on the growth cycle includes: Conducting numerical analysis on the stability parameters of the strains, establishing an evaluation matrix including stability index, bacterial wilt resistance index and yield index, and constructing a comprehensive evaluation system for breeding materials; Quantitatively calculate each indicator in the comprehensive evaluation system of breeding materials according to weights to form a multi-dimensional evaluation score and establish a comprehensive scoring data set; Performing clustering processing on the comprehensive score data set, dividing the breeding materials into different grades according to the score distribution, and forming a graded evaluation result; A screening threshold is set according to the grading evaluation results, and materials above the threshold are included in the excellent strain candidate set to construct an excellent material database; Analyze the growth cycle characteristics of the strains in the excellent material database, evaluate the bacterial wilt resistance and growth status throughout the growth period, and form a strain characteristic profile; Based on the strain characteristic profile, excellent strains are screened, and the growth cycle performance and resistance to bacterial wilt are comprehensively considered to obtain a breeding plan for resistance to bacterial wilt based on the growth cycle.

8. The method for analyzing peanut bacterial wilt resistance breeding strategies based on growth cycles according to claim 7, characterized in that: The various indicators in the comprehensive evaluation system of breeding materials are quantified and calculated according to weights to form a multi-dimensional evaluation score and establish a comprehensive scoring data set, including: Extracting stability index data from the comprehensive evaluation system of breeding materials, including genotype effect value, environmental effect value and interaction effect value in strain stability parameters, and constructing a stability scoring matrix; Extracting bacterial wilt resistance index data from the comprehensive evaluation system of breeding materials, including disease incidence, disease index, lesion size and disease progression rate, and constructing a bacterial wilt resistance scoring matrix; Extracting yield index data from the comprehensive evaluation system of breeding materials, including single plant yield, pod number, fruit setting rate and 100-fruit weight, and constructing a yield scoring matrix; Setting weight coefficients for the stability scoring matrix, the bacterial wilt resistance scoring matrix, and the yield scoring matrix to construct a weight distribution scheme; Perform weighted calculation on the scoring matrix based on the weight distribution scheme, integrate the scores of various indicators, and form a comprehensive score table; The comprehensive score table is normalized to unify various indicators to the same scale to obtain a comprehensive score data set.

9. The method for analyzing peanut bacterial wilt resistance breeding strategies based on growth cycles according to claim 8, characterized in that: The method extracts bacterial wilt resistance index data from the comprehensive evaluation system of breeding materials, including disease incidence, disease index, lesion size and disease progression rate, and constructs a bacterial wilt resistance scoring matrix, including: The disease incidence data were divided into seedling incidence, vegetative growth period incidence and reproductive growth period incidence according to different growth periods, and the incidence time series matrix was constructed. The disease index data are classified and counted according to the lesion area, wilting degree and affected parts, and a disease classification data table is established; Extract the time series changes of lesion length, width and area from the lesion size data to form a lesion development data set; Track the disease progression rate data in time series, calculate the lesion expansion speed and disease severity per unit time, and obtain the disease dynamic parameters; Integrate the incidence time series matrix, disease grading data table, lesion development data set and disease dynamic parameters to form a bacterial wilt resistance characteristic database; Based on the various indicators in the bacterial wilt resistance characteristic database, weighted processing is performed, and a comprehensive score is calculated by combining weight coefficients to construct a bacterial wilt resistance scoring matrix.

10. A peanut bacterial wilt resistance breeding strategy analysis system based on growth cycle, used to implement the peanut bacterial wilt resistance breeding strategy analysis method based on growth cycle as described in any one of claims 1 to 9, characterized in that: The peanut bacterial wilt resistance breeding strategy analysis system based on the growth cycle includes: The acquisition module is used to continuously collect the morphological characteristics, physiological indicators and disease symptoms of peanuts during their growth period through the phenotypic monitoring system, divide the data according to the key growth period through time series segmentation, construct a time series feature vector, and obtain a growth cycle feature library; A comparison module is used to compare and analyze the phenotypic trajectories of healthy plants and diseased plants according to the growth cycle feature library, establish early warning indicators through differential feature extraction, and obtain an early warning model for resistance to bacterial wilt; A monitoring module is used to monitor and analyze the growth phenotype data of breeding materials in real time according to the bacterial wilt resistance early warning model, determine potential susceptible individuals through early warning thresholds, and obtain risk early warning results; A tracking module is used to track plants with different resistance levels throughout their growth period based on the risk warning results, establish a resistance scoring system by comparing phenotypic characteristics during key periods, and obtain a dynamic resistance evaluation scheme; An analysis module, for collecting multi-point test data for environmental response analysis according to the dynamic resistance evaluation scheme, establishing adaptability indicators through plant-environment interaction evaluation, and obtaining strain stability parameters; The grading module is used to comprehensively evaluate and grade the breeding materials according to the strain stability parameters, determine the excellent strains through a grading screening system, and obtain a breeding plan for resistance to bacterial wilt based on the growth cycle.

Images