Phenomics analysis-based macrobrachium rosenbergii chela ratio prediction and breeding method

Abstract

The invention discloses a macrobrachium rosenbergii cherax ratio prediction and breeding method based on phenotypic omics analysis, which comprises the following steps: selecting a plurality of macrobrachium rosenbergii for multi-generation breeding, recording pedigree data of each individual, and constructing a macrobrachium rosenbergii breeding database; the method comprises the following steps: extracting phenotypic characteristics of macrobrachium at different culture stages, analyzing a rule between a growth index and a chelate ratio, carrying out heritability evaluation, and constructing a macrobrachium phenotypic omics network; the growth path of macrobrachium rosenbergii is analyzed, abnormal events in the breeding period are recognized, and a macrobrachium rosenbergii growth situation network is generated; constructing a macrobrachium phenotype knowledge map, and constructing a macrobrachium cherax ratio prediction model; and performing data acquisition on a macrobrachium rosenbergii population to be bred to obtain early phenotypic data, performing breeding decision, and performing breeding path recommendation on bred individuals by using the macrobrachium rosenbergii phenotypic knowledge graph. Meanwhile, whether the breeding path is deviated or not in the breeding process is analyzed, and a correction scheme is formulated for pushing, so that the breeding period of the macrobrachium is shortened, and the breeding accuracy and efficiency are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of Macrobrachium rosenbergii breeding technology, and particularly relates to a Macrobrachium rosenbergii chelicera ratio prediction and breeding method based on phenomics analysis. BACKGROUND

[0002] As a large freshwater prawn with important economic value, Macrobrachium rosenbergii has been widely developed in the global aquaculture industry. However, in the actual large-scale breeding process, a long-standing and difficult-to-eliminate technical bottleneck seriously restricts the further improvement of its yield and the sustainable development of the industry, that is, the fierce mutual killing behavior among individual Macrobrachium rosenbergii. This behavior directly leads to a significant decrease in survival rate, an increase in feed conversion ratio, and ultimately causes huge economic losses. Studies have shown that the chelae (second pereiopods) of male Macrobrachium rosenbergii are abnormally developed, and the length can even reach twice the body length. This morphological characteristic is positively correlated with the strong territoriality and aggressiveness of the behavior. In other words, the more developed the chelae, the stronger the aggressiveness, and the greater the harm to the same species in a high-density breeding environment, which is the main cause of mutual killing.

[0003] Traditional breeding improvement methods usually rely on the experience of breeding personnel, and artificial selection is performed according to a few intuitive phenotypic traits such as the size and weight of adult prawns in the later breeding period. This method has fundamental defects in reducing the chelicera ratio (the ratio of chelae length to body length), a complex trait. First, the breeding cycle is prolonged, and accurate measurement and selection must be performed after the prawns grow to sexual maturity and the chelae are fully developed, which greatly prolongs the generation interval, seriously slows down the breeding process, and increases the breeding cost. Second, the traditional method has low throughput and strong subjectivity, and it is difficult to accurately and efficiently measure large-scale populations, and it is also impossible to predict the future chelae growth and development trend at an early stage, resulting in low accuracy and efficiency of selection. Therefore, there is an urgent need for an innovative technical solution that can achieve early, rapid, and accurate breeding of low-chelicera-ratio and low-aggressiveness Macrobrachium rosenbergii new strains.

[0004] With the development of modern information technology, phenomics, big data analysis, and machine learning technology provide new ideas for solving this problem. Through automatic, high-throughput image acquisition and processing technology, massive morphological phenotype data of prawn bodies during the entire growth period can be obtained without damage. Further, by combining pedigree information, the heritability of the chelicera ratio trait can be accurately evaluated using quantitative genetics methods to confirm the potential for genetic improvement. Therefore, the present application provides a Macrobrachium rosenbergii chelicera ratio prediction and breeding method based on phenomics, thereby shortening the breeding cycle of Macrobrachium rosenbergii, improving the accuracy and efficiency of breeding, and assisting in the breeding of prawn individuals. SUMMARY

[0005] The present application overcomes the defects of the prior art and provides a Macrobrachium rosenbergii chela ratio prediction and breeding method based on phenomics analysis.

[0006] To achieve the above-mentioned purpose, the first aspect of the present application provides a Macrobrachium rosenbergii chela ratio prediction and breeding method based on phenomics analysis, comprising: Selecting a plurality of Macrobrachium rosenbergii for multi-generation breeding, recording pedigree data of each individual during the breeding process, and collecting phenotype data of the Macrobrachium rosenbergii at different breeding stages and breeding control data to construct a Macrobrachium rosenbergii breeding database; Extracting Macrobrachium rosenbergii phenotype characteristics at different breeding stages through the Macrobrachium rosenbergii breeding database, analyzing the relationship between individual growth indicators and chela ratio, and performing genetic force evaluation to construct a Macrobrachium rosenbergii phenomics network; Analyzing the growth path of Macrobrachium rosenbergii during breeding using the Macrobrachium rosenbergii breeding database, identifying abnormal events during breeding and associating corresponding disposal schemes, and generating a Macrobrachium rosenbergii growth trend network; Constructing a Macrobrachium rosenbergii phenotype knowledge graph in combination with the Macrobrachium rosenbergii phenomics network and the Macrobrachium rosenbergii growth trend network, and constructing a Macrobrachium rosenbergii chela ratio prediction model using the Macrobrachium rosenbergii phenotype knowledge graph; Collecting early phenotype data of a Macrobrachium rosenbergii population to be bred, making breeding decisions through the Macrobrachium rosenbergii chela ratio prediction model, and recommending breeding paths for breeding individuals using the Macrobrachium rosenbergii phenotype knowledge graph; Obtaining multi-modal monitoring data of breeding individuals during the breeding process, analyzing whether the actual growth trajectory or environmental parameters deviate from the expected path, and if there is a deviation, developing a correction scheme for pushing.

[0007] In the present scheme, the selection of a plurality of Macrobrachium rosenbergii for multi-generation breeding, the recording of pedigree data of each individual during the breeding process, and the collection of Macrobrachium rosenbergii phenotype data at different breeding stages and breeding control data to construct a Macrobrachium rosenbergii breeding database specifically comprises: Selecting a plurality of Macrobrachium rosenbergii with chela ratios meeting the expected threshold in the existing breeding pond as a one-generation breeding population, giving each Macrobrachium rosenbergii parent and offspring entering the pond a unique electronic identifier, and constructing a Macrobrachium rosenbergii breeding pedigree through multi-generation breeding; During the breeding process, the array of breeding monitoring sensors periodically collects image data of Macrobrachium rosenbergii individuals at different growth stages, Macrobrachium rosenbergii breeding pond environment data, and Macrobrachium rosenbergii control data to generate an original multi-modal data set; The obtained original multi-modal data set is pre-processed, the collected individual image data of macrobrachium is introduced into an image segmentation model built based on a convolutional neural network to obtain individual segmentation images of the macrobrachium, morphological phenotype data of the individual is calculated according to a preset pixel-actual size conversion coefficient, and is bound with a corresponding individual ID and a collection time stamp to generate a macrobrachium phenotype data set; For the macrobrachium culture pond environment data and the macrobrachium control data, data cleaning is performed by using an outlier monitoring algorithm, abnormal measurement values are identified and removed, environmental data fluctuation characteristics of time periods before and after event points are extracted according to event time stamps recorded in the culture log, and a control label is generated for each event, the control label including control measures, environmental parameter changes before and after the control, and macrobrachium state data, and finally a culture control data set with event semantics is generated; A macrobrachium culture database is constructed in combination with a macrobrachium culture pedigree, the macrobrachium phenotype data set and the culture control data set.

[0008] In the scheme, the macrobrachium phenotype characteristics of different culture stages are extracted from the macrobrachium culture database, and the relationship between individual growth indicators and chelae-body ratios is analyzed, specifically including: The macrobrachium culture database is obtained, complete time sequence phenotype data of all macrobrachium individuals at each time point is extracted from the macrobrachium culture database, a macrobrachium phenotype measurement sequence is generated, and a fixed time length sliding window is used to segment the macrobrachium phenotype measurement sequence to generate a plurality of macrobrachium phenotype measurement sub-sequences; First-order differences of adjacent time points of each macrobrachium phenotype measurement sub-sequence are calculated as instantaneous growth rate characteristics, and differences of the first-order difference sequence are calculated as second-order differences to represent growth acceleration; Meanwhile, a complete growth curve of each phenotype indicator is subjected to cubic spline interpolation or polynomial fitting, coefficients of the fitting function are extracted as shape characteristics describing the overall growth pattern, and statistical characteristics of each sequence are calculated, including mean, variance, skewness and kurtosis, and finally a macrobrachium phenotype derived characteristic sequence is obtained; The macrobrachium phenotype measurement sequence and the macrobrachium phenotype derived characteristic sequence are integrated to construct a multi-modal feature set representing the growth dynamic pattern of the individual, and correlation analysis is performed in combination with the finally determined chelae-body ratio of each macrobrachium individual, the maximum mutual information coefficient and the Pearson correlation coefficient between each feature and the chelae-body ratio are calculated; Subsequently, a random forest regression model is called, the finally determined chelae-body ratio is taken as a target variable, the macrobrachium phenotype measurement characteristics and the macrobrachium phenotype extended characteristics are taken as decision variables, the importance score of each feature is calculated through ten-fold cross-validation, the key growth characteristics and weight coefficients significantly related to the chelae-body ratio are analyzed, and a Gaussian process regression model is established through the screened key growth characteristics to fit the nonlinear mapping relationship between the phenotype characteristics and the target trait, and finally a phenotype characteristic-chelae-body ratio correlation rule set is generated.

[0009] In the scheme, the genetic evaluation is performed, and a macrobrachium nipponense phenomics network is constructed, specifically comprising: Obtaining a set of phenotype feature-chelicera ratio association rules, extracting key growth characteristics and weight coefficients significantly related to chelicera ratio from the set of phenotype feature-chelicera ratio association rules to construct a fixed effect design matrix, extracting a pedigree table from a macrobrachium nipponense breeding database, and using the pedigree table to construct a kinship matrix of all macrobrachium nipponense individuals; Based on the fixed effect design matrix and the kinship matrix, the additive genetic effect of the macrobrachium nipponense individual is taken as a random effect, and the kinship matrix is defined as the variance-covariance matrix prior of the random effect, and a mixed linear model based on an animal model is constructed; The mixed linear model is fitted by using a restricted maximum likelihood method, the variance components are iteratively solved by an EM algorithm, the additive genetic variance component and the residual variance component of the chelicera ratio trait are obtained, the genetic force estimate value is obtained by calculating the ratio of the additive genetic variance to the phenotypic value, and the breeding value of each individual is solved by a BLUP method; Based on the set of phenotype feature-chelicera ratio association rules, the key growth characteristics are extracted as network nodes, the dynamic time warping algorithm is used to calculate the phenotype distance matrix between individuals to represent the asynchronous growth trajectory of the macrobrachium nipponense individuals, the graphical lasso algorithm is used to estimate the conditional independence relationship between the characteristics, and a sparse Gaussian graph model is constructed; The sparse Gaussian graph model is defined as the basic topological structure of the phenomics network, wherein the node represents the phenotype characteristic, and the edge weight represents the conditional correlation coefficient between the characteristics; the association strength of each phenotype characteristic with the chelicera ratio is extracted as the initial attribute of the node through the set of phenotype feature-chelicera ratio association rules, the genetic force estimate value is taken as the global attribute of the network, and the individual breeding value is taken as the additional attribute of the corresponding individual phenotype node, to generate an initial phenomics network; The Node2Vec algorithm is used to learn the low-dimensional vector representation of the nodes of the initial phenomics network, the transition probability between the nodes is defined, the node sequence is generated by biased random walk for node embedding vector learning, and finally the high-dimensional node embedding is mapped to a two-dimensional space by using the t-SNE dimension reduction algorithm, to generate a macrobrachium nipponense phenomics network.

[0010] In the scheme, the growth path of the macrobrachium nipponense in the breeding process is analyzed by using the macrobrachium nipponense breeding database, abnormal events during the breeding period are identified and corresponding treatment schemes are associated, and a macrobrachium growth trend network is generated, specifically comprising: Obtaining a macrobrachium culture database, extracting time series phenotype measurement data and corresponding environmental parameter sequences of all macrobrachium individuals from the macrobrachium culture database, and after preprocessing, using a nonlinear mixed effect model to fit the standard growth curve of macrobrachium rosenbergii by taking individuals as random effects, and establishing an individual growth curve model describing the normal growth pattern; Based on the established individual growth curve model, the residual sequence between the actual measured value and the model predicted value of each individual is calculated, and wavelet transform analysis is performed on the residual sequence to extract time-frequency domain features and generate a macrobrachium individual growth trajectory set; Abnormal events in the growth process are detected by combining environmental parameter sequences through isolation forest algorithm, and the occurrence time, event characteristics, duration and abnormal intensity index of abnormal events are recorded to generate an abnormal growth event feature set; According to the abnormal growth event feature set, the identified abnormal event points are time-stamped matched with the control log records in the culture control data set, if there is an artificial recorded control operation for a certain abnormal event point, the corresponding abnormal event and control measure are associated to form an abnormal event-disposal scheme association pair, otherwise it is an environmental self-fluctuation event, and an abnormal event-self fluctuation association pair is generated, which is marked as an event without disposal; Based on the obtained association pair, an abnormal-disposal mapping knowledge base is constructed in the order of abnormal event features, environmental parameter features and disposal scheme, including abnormal event type, environmental parameter feature, disposal measure taken and effect after disposal; Through the macrobrachium individual growth trajectory set, the growth trajectory features of each macrobrachium individual are extracted to generate a plurality of growth trajectory feature sequences with time series attributes, based on the abnormal-disposal mapping knowledge base, based on the identification of the corresponding macrobrachium individual of the abnormal event and the time stamp of the occurrence of the abnormal event, the corresponding growth trajectory feature sequence is merged to generate a growth trend sequence representing the growth trend of each macrobrachium individual in the culture process; The cosine similarity and Mahalanobis distance between each growth trend sequence are calculated respectively, and weighted average and normalization processing are performed to generate a sequence merging index, which is compared with a preset threshold, if it is greater than the preset threshold, the difference between the two growth trend sequences is calculated to identify the difference sequence segment, and a new sequence branch is generated in the difference sequence segment to merge the sequences, and finally a macrobrachium growth trend network is generated through repeated iteration and merging steps.

[0011] In the scheme, the macrobrachium phenotype knowledge graph is constructed by combining the macrobrachium phenomics network and the macrobrachium growth trend network, and the macrobrachium chelicera ratio prediction model is constructed by using the macrobrachium phenotype knowledge graph, which specifically includes: Acquire a macrobrachium rosenbergii phenotype network and a macrobrachium rosenbergii growth trend network, extract all phenotype feature nodes and node attribute information from the macrobrachium rosenbergii phenotype network, including the correlation strength of growth characteristics and chelae ratio, genetic force estimation value and individual breeding value, and extract abnormal event nodes, treatment scheme nodes and their correlation from the macrobrachium rosenbergii growth trend network; Based on the extracted phenotype feature nodes, generate a phenotype node sequence main shaft with the phenotype change time sequence of the growth process of the macrobrachium rosenbergii individual as the node connection sequence, which represents all phenotype change patterns of the macrobrachium rosenbergii individual in the growth process; connect the abnormal event nodes and the treatment scheme nodes based on the correlation between the abnormal event nodes and the treatment scheme nodes as the connection basis to generate node connection edges, and generate a node sequence sub-shaft; Merge the node sequence sub-shaft with the node sequence main shaft according to the trigger event stamp of the abnormal event as the index to obtain a plurality of tree-shaped node sequences, which represent the phenotype change patterns and growth trend of the macrobrachium rosenbergii individual in the cultivation process, and finally form a macrobrachium rosenbergii phenotype knowledge graph based on all tree-shaped node sequences; Perform low-dimensional vector representation learning on the entities and relationships of the macrobrachium rosenbergii phenotype knowledge graph through a TransE knowledge identification learning algorithm and construct a training data set, build a macrobrachium rosenbergii chelae ratio prediction model based on a graph neural network as the basic architecture, and train the model using the training data set, optimize the hyperparameters through cross-validation and grid search optimization, and finally obtain a macrobrachium rosenbergii chelae ratio prediction model that meets the expectations.

[0012] In this scheme, the early phenotype data of the to-be-selected breeding macrobrachium rosenbergii population is acquired through data acquisition, breeding decisions are made through the macrobrachium rosenbergii chelae ratio prediction model, and the breeding path of the selected individual is recommended using the macrobrachium rosenbergii phenotype knowledge graph, which specifically includes: When the to-be-selected breeding macrobrachium rosenbergii population is cultivated, the morphological phenotype data of each individual is acquired through the deployed automatic image acquisition system, and the cultivation environment parameters are simultaneously monitored and acquired to obtain cultivation environment monitoring data; The acquired early phenotype data and cultivation environment monitoring data are preprocessed and standardized to form early monitoring feature vectors of the to-be-selected breeding individuals; the early monitoring feature vectors are input into the trained macrobrachium rosenbergii chelae ratio prediction model, the prediction value of the chelae ratio of the individual after adulthood and the corresponding prediction confidence interval are output through multi-layer nonlinear transformation and feature extraction inside the model; Based on the chelae ratio prediction results of all to-be-selected breeding individuals, the chelae ratio prediction values are sorted, the prediction confidence interval and the preset selection intensity threshold are combined, and the macrobrachium rosenbergii individuals with chelae ratio meeting the preset expectations are selected as candidate parents to generate a macrobrachium rosenbergii breeding recommendation table; According to the macrobrachium recommended breeding table, combined with the macrobrachium phenotype knowledge graph for breeding path analysis, taking the early monitoring feature vector and pedigree information of the selected individual as the query condition, the historical individual nodes with similar phenotype characteristics and genetic background are retrieved in the knowledge graph, the best environmental parameter range experienced by the historical individual nodes in the complete breeding cycle, the successful control measure sequence and the finally reached growth performance index are obtained, and the similar breeding case is generated; Based on the retrieved similar breeding case, the case-based reasoning technology is used to generate the individualized breeding path recommendation scheme of the target recommended breeding individual, including the environmental parameter control target, feeding strategy, density management requirement and expected abnormal events and countermeasures in different growth stages; The macrobrachium breeding recommendation table is associated with the individualized breeding path recommendation scheme to generate the final macrobrachium breeding decision report for pushing.

[0013] In the scheme, the multi-modal monitoring data of the breeding individual in the breeding process is obtained, and whether the actual growth trajectory or the environmental parameter deviates from the expected path is analyzed, and if there is a deviation, a correction scheme is developed for pushing, which specifically includes: Real-time acquisition of multi-modal monitoring data of the breeding individual in the breeding process, including time-series environmental parameters collected by a sensor array and individual phenotype measurement data periodically acquired by an image acquisition system; The multi-modal monitoring data is compared with the preset breeding path expected value recommended for the individual in the macrobrachium phenotype knowledge graph, the dynamic time warping algorithm is used to calculate the similarity distance between the actual growth sequence and the expected growth sequence, and the deviation threshold is set based on the statistical process control method; When it is detected that the actual growth trajectory or the environmental parameter continuously deviates from the expected path and exceeds the deviation threshold, the early warning mechanism is triggered; according to the deviation characteristics, the abnormal-disposal mapping knowledge base in the macrobrachium growth trend network is queried, and the historical similar deviation mode and the corresponding effective disposal scheme are matched; at the same time, the genetic background information and the phenotype characteristic association rule of the individual in the macrobrachium phenotype knowledge graph are retrieved, and the potential influence of the specific disposal measure on the genetic strain is evaluated; Comprehensive historical disposal scheme effect and individual genetic characteristics, generate individualized correction scheme for current deviation, the correction scheme includes environmental parameter adjustment strategy, feeding management measure adjustment suggestion and expected recovery trajectory; the correction scheme is pushed to the breeding management system and the early warning prompt is triggered at the same time, guiding manual or automatic execution of control operation; After the implementation of the correction scheme, the growth response of the individual and the change of the environmental parameters are continuously monitored, the actual recovery trajectory is compared with the expected recovery trajectory, and the implementation effect of the correction scheme is evaluated; the present deviation event, the correction measures taken and the final effect are fed back and updated to the marsh shrimp growth trend network and the marsh shrimp phenotype knowledge graph as new knowledge samples, so as to realize the continuous optimization and learning of the knowledge base.

[0014] Another aspect of the present application provides a computer-readable storage medium comprising a macrobrachium rosenbergii chela ratio prediction and breeding method program based on phenomics analysis, wherein the macrobrachium rosenbergii chela ratio prediction and breeding method program based on phenomics analysis is executed by a processor to realize the steps of the macrobrachium rosenbergii chela ratio prediction and breeding method based on phenomics analysis according to any one of the above aspects. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the embodiments or examples of the present application, the drawings needed in the embodiment or example description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained according to the drawings shown by those skilled in the art without creating any creative labor.

[0016] Figure 1 A first method flowchart of a macrobrachium rosenbergii chela ratio prediction and breeding method based on phenomics analysis provided by an embodiment of the present application; Figure 2 A second method flowchart of a macrobrachium rosenbergii chela ratio prediction and breeding method based on phenomics analysis provided by an embodiment of the present application; Figure 3 A third method flowchart of a macrobrachium rosenbergii chela ratio prediction and breeding method based on phenomics analysis provided by an embodiment of the present application; The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0017] In order to more clearly illustrate the technical solutions in the embodiments or examples of the present application, the drawings needed in the embodiment or example description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained according to the drawings shown by those skilled in the art without creating any creative labor.

[0018] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, however, the present application can also be implemented in other ways different from those described herein, therefore, the protection scope of the present application is not limited by the specific embodiments disclosed below.

[0019] Figure 1 This is a flowchart of the first method for predicting and breeding the claw-body ratio of Macrobrachium rosenbergii based on phenomics analysis, as provided in an embodiment of the present invention. like Figure 1 As shown, this invention provides a first method flowchart for predicting and breeding the claw-to-body ratio of Macrobrachium rosenbergii based on phenomics analysis, including: S102, Select a number of giant freshwater prawns for multi-generation farming, record the pedigree data of each individual during the farming process, and collect phenotypic data and farming management data of giant freshwater prawns at different farming stages to construct a giant freshwater prawn farming database. S104. Phenotypic characteristics of prawns at different farming stages are extracted from the prawn farming database, the relationship between individual growth indicators and chelicerate ratio is analyzed, heritability is assessed, and a prawn phenomics network is constructed. S106, Analyze the growth path of giant freshwater prawns during the farming process using the prawn farming database, identify abnormal events during the farming period and associate them with corresponding treatment plans, and generate a prawn growth status network. S108, Combine the above-mentioned prawn phenomics network and prawn growth status network to construct a prawn phenotypic knowledge graph, and use the prawn phenotypic knowledge graph to construct a prawn claw-body ratio prediction model. S110, Collect early phenotypic data of the giant freshwater prawn population to be bred, make bred selection decisions through the giant freshwater prawn claw-to-body ratio prediction model, and recommend farming paths for selected individuals using the giant freshwater prawn phenotypic knowledge graph. S112: Obtain multimodal monitoring data of selected individuals during the breeding process, analyze whether the actual growth trajectory or environmental parameters deviate from the expected path, and if there is a deviation, formulate a correction plan and push it out.

[0020] Furthermore, in a preferred embodiment of the present invention, the step of selecting several giant freshwater prawns for multi-generational culture, recording the pedigree data of each individual during the culture process, and collecting prawn phenotypic data and culture management data at different culture stages to construct a giant freshwater prawn culture database specifically includes: Several giant freshwater prawns with a chelicer-to-body ratio that meet the desired threshold are selected from existing culture ponds as a first-generation culture population. Each giant freshwater prawn parent and offspring is given a unique electronic identifier. A culture pedigree of giant freshwater prawns is constructed through multiple generations of culture. During the aquaculture process, an array of aquaculture monitoring sensors is deployed to periodically collect individual image data of prawns at different growth stages, environmental data of prawn farming ponds, and prawn management data to generate a raw multimodal dataset. The obtained original multi-modal data set is preprocessed, the collected individual image data of macrobrachium is introduced into an image segmentation model built based on a convolutional neural network framework to obtain individual segmentation images of macrobrachium, morphological phenotype data of the individual is calculated according to a preset pixel-actual size conversion coefficient, and the individual ID and the collection time stamp are bound to generate a macrobrachium phenotype data set; For the macrobrachium breeding pond environment data and the macrobrachium control data, data cleaning is performed by using an outlier monitoring algorithm, abnormal measurement values are identified and removed, environmental data fluctuation characteristics of time periods before and after an event point are extracted according to an event time stamp recorded in a breeding log, and a control label is generated for each event, the control label including a control measure, environmental parameter changes before and after the control, and macrobrachium state data, and finally a breeding control data set with event semantics is generated. The macrobrachium breeding database is constructed in combination with the macrobrachium breeding pedigree, the macrobrachium phenotype data set and the breeding control data set.

[0021] It should be noted that in the existing breeding pond, individuals with potential excellent traits are selected according to a clear phenotype standard, i.e., the chelae-body ratio (cheliped length to body length ratio) meets a pre-set expected threshold, to form a core basic breeding population. To ensure the uniqueness of subsequent data tracing and individual identification, each selected parent and its offspring individual is assigned a unique electronic identifier. During the breeding process across multiple generations, a set of integrated monitoring sensor arrays are systematically arranged in the breeding environment. The arrays periodically capture multi-modal raw data: individual images of macrobrachium at different growth stages (such as larval stage, juvenile stage, adult stage) are obtained through high-definition industrial cameras; environmental parameters of the breeding pond are continuously recorded through water quality sensors (such as temperature, pH, dissolved oxygen, ammonia nitrogen sensors); and all manual intervention operations (such as feeding, water changing, pond dividing, medication, etc.) are recorded through a digitalized log system, which together constitute the original multi-modal data set. The preprocessing of the above raw data set is a key link to ensure data quality and availability. The individual image data is introduced into an image segmentation model pre-trained based on a convolutional neural network (such as U-Net or Mask R-CNN architecture), which can accurately identify and segment the key morphological parts such as shrimp body outline and cheliped in the image. After segmentation, accurate morphological phenotype data such as body length and cheliped length are automatically calculated according to the pixel-actual size conversion coefficient determined in advance through a calibration board. These quantitative data are then strictly bound with the corresponding individual unique ID and collection time stamp, thereby generating a structured macrobrachium phenotype data set with time sequence characteristics.

[0022] For the continuous monitoring of environmental data and artificial record management data, an outlier monitoring algorithm (such as Isolation Forest or Z-Score algorithm) is used for data cleaning to identify and eliminate abnormal measurement values caused by sensor transient failure or human recording error, ensuring the reliability of the data sequence. Further, according to the time stamp of the management event recorded in the breeding log, the environmental data fluctuation characteristics (such as temperature change curve, dissolved oxygen recovery rate) within a certain time window before and after the event point are extracted, and a structured management label is generated for each event. The label not only records the management measures itself, but also encapsulates the dynamic changes of environmental parameters before and after the management and the observed Macrobrachium rosenbergii population state data (such as feeding behavior, activity) at that time, ultimately generating a breeding management data set rich in contextual semantics. Finally, through data fusion technology, the above three types of data processed—revealing genetic association Macrobrachium rosenbergii breeding pedigree, recording individual growth history of Macrobrachium rosenbergii phenotype data set, and describing environment and management intervention breeding management data set—are integrated to build a spatiotemporally synchronized, structurally standardized, and informationally complete Macrobrachium rosenbergii breeding database.

[0023] Further, in a preferred embodiment of the present application, the extraction of Macrobrachium rosenbergii phenotype characteristics at different breeding stages from the Macrobrachium rosenbergii breeding database and the analysis of the relationship between individual growth indicators and chelae body ratio specifically include: Obtain the Macrobrachium rosenbergii breeding database, extract the complete time series phenotype data of all Macrobrachium rosenbergii individuals at each time point from the Macrobrachium rosenbergii breeding database, generate Macrobrachium rosenbergii phenotype measurement sequences, and use a sliding window of fixed time length to segment the Macrobrachium rosenbergii phenotype measurement sequences to generate several Macrobrachium rosenbergii phenotype measurement sub-sequences; Calculate the first-order difference of adjacent time points of each Macrobrachium rosenbergii phenotype measurement sub-sequence as the instantaneous growth rate feature, and calculate the difference of the first-order difference sequence as the second-order difference to represent the growth acceleration; At the same time, perform cubic spline interpolation or polynomial fitting on the complete growth curve of each phenotype indicator, extract the coefficients of the fitting function as shape features to describe the overall growth pattern, and calculate the statistical features of each sequence, including mean, variance, skewness and kurtosis, to finally obtain the Macrobrachium rosenbergii phenotype derived feature sequence; Integrate the Macrobrachium rosenbergii phenotype measurement sequence and the Macrobrachium rosenbergii phenotype derived feature sequence to construct a multi-modal feature set representing the individual growth dynamic pattern, and perform correlation analysis combined with the finally determined chelae body ratio of each Macrobrachium rosenbergii individual to calculate the maximum mutual information coefficient and Pearson correlation coefficient between each feature and the chelae body ratio; Then, the random forest regression model is called to take the finally determined chelicera body ratio as the target variable and the macrobrachium phenotypic measurement features and the macrobrachium phenotypic derived features as the decision variables, the importance score of each feature is calculated through ten-fold cross-validation, the key growth features and the weight coefficients significantly related to the chelicera body ratio are analyzed, and the Gaussian process regression model is established based on the screened key growth features to fit the nonlinear mapping relationship between the phenotypic features and the target traits, and finally the phenotypic feature-chelicera body ratio association rule set is generated.

[0024] It should be noted that the complete phenotypic measurement data of all individuals recorded in chronological order during the entire culture cycle is extracted from the constructed macrobrachium culture database to form the macrobrachium phenotypic measurement sequence of each individual. In order to capture the dynamic characteristics at different growth stages, a sliding window with a fixed time length is used to segment and cut the sequence, thereby generating a series of macrobrachium phenotypic measurement sub-sequences that can reflect local growth characteristics. For each phenotypic measurement sub-sequence, the first-order difference of the measurement values at adjacent time points is calculated to quantify the instantaneous growth rate, and then the first-order difference sequence is differentiated again to obtain the second-order difference to represent the acceleration change pattern of growth. At the same time, for the complete growth trajectory of each phenotypic index (such as body length and cheliped length), a cubic spline interpolation is used for smooth fitting or a polynomial function is used for trend approximation, and the coefficients of the fitting function are extracted as shape features to describe the overall growth morphology. In addition, the statistical features of each sequence are calculated, including the mean reflecting the average level, the variance representing the fluctuation amplitude, the skewness indicating the symmetry of the distribution, and the kurtosis measuring the steepness of the distribution, which together constitute the macrobrachium phenotypic derived feature sequence. The original measurement sequence and the derived feature sequence are integrated to construct a multi-modal feature set that can comprehensively represent the growth dynamic pattern of the individual. The feature set is associated with the real value of the finally determined chelicera body ratio of each individual, the maximum mutual information coefficient is used to capture the nonlinear correlation strength between the features and the target traits, and the Pearson correlation coefficient is calculated to measure the linear correlation degree, thereby preliminarily screening the feature indexes closely related to the chelicera body ratio. Based on the above analysis results, the random forest regression model is called for in-depth feature screening and modeling. The final chelicera body ratio is taken as the target variable, and all phenotypic measurement features and derived features are taken as the decision variables, the importance score of each feature is evaluated through ten-fold cross-validation, and the key growth features and the corresponding weight coefficients significantly related to the chelicera body ratio are identified. Finally, based on the screened key features, the Gaussian process regression model is established to accurately fit the complex nonlinear mapping relationship between the phenotypic features and the target traits, which can provide uncertainty estimation of the prediction results, and finally generate the phenotypic feature-chelicera body ratio association rule set containing the feature-trait association rules and their confidence levels, providing quantitative basis for subsequent genetic evaluation and breeding decision.

[0025] Further, in a preferred embodiment of the present application, the genetic force evaluation and the construction of the macrobrachium nipponense phenomics network specifically include: Obtaining a set of phenotype feature-chelicera ratio association rules, extracting key growth characteristics and weight coefficients significantly related to the chelicera ratio from the set of phenotype feature-chelicera ratio association rules to construct a fixed effect design matrix, extracting a pedigree table from the macrobrachium nipponense breeding database, and using the pedigree table to construct a kinship matrix of all macrobrachium nipponense individuals; Based on the fixed effect design matrix and the kinship matrix, the additive genetic effect of the macrobrachium nipponense individual is taken as a random effect, and the kinship matrix is defined as the variance-covariance matrix prior of the random effect to construct a mixed linear model based on the animal model; The mixed linear model is fitted by using the restricted maximum likelihood method, the variance components are solved by the EM algorithm, the additive genetic variance component and the residual variance component of the chelicera ratio trait are obtained, the genetic force estimate is obtained by calculating the ratio of the additive genetic variance to the phenotypic value, and the breeding value of each individual is solved by the BLUP method; Based on the set of phenotype feature-chelicera ratio association rules, the key growth characteristics are extracted as network nodes, the dynamic time warping algorithm is used to calculate the phenotype distance matrix between individuals to represent the asynchronous growth trajectory of the macrobrachium nipponense individuals, the graphical lasso algorithm is used to estimate the conditional independence relationship between the characteristics, and a sparse Gaussian graph model is constructed; The sparse Gaussian graph model is defined as the basic topology structure of the phenomics network, wherein the nodes represent the phenotype characteristics, and the edge weight represents the conditional correlation coefficient between the characteristics; the association strength of each phenotype characteristic with the chelicera ratio is extracted as the initial attribute of the node through the set of phenotype feature-chelicera ratio association rules, the genetic force estimate is taken as the global attribute of the network, and the individual breeding value is taken as the additional attribute of the corresponding individual phenotype node to generate an initial phenomics network; The Node2Vec algorithm is used to learn the low-dimensional vector representation of the nodes of the initial phenomics network, the transition probability between the nodes is defined, the node sequence is generated by biased random walk to learn the node embedding vector, and finally the t-SNE dimension reduction algorithm is used to map the high-dimensional node embedding to a two-dimensional space to generate the macrobrachium nipponense phenomics network.

[0026] It should be noted that the obtained phenotype characteristics-chelicera ratio association rule set is used to extract key growth characteristics and their weight coefficients significantly related to the chelicera ratio, and a fixed effect design matrix for genetic evaluation is constructed. At the same time, the complete pedigree information table is extracted from the macrobrachium rosenbergii breeding database, and the kinship coefficient calculation rule is used to construct the kinship matrix of all individuals based on the pedigree data, which quantifies the genetic similarity between individuals in the population. Based on the fixed effect design matrix and the kinship matrix, a mixed linear model based on the animal model is constructed. In this model, the screened key phenotype characteristics are used as fixed effects, the additive genetic effects of individuals are used as random effects, and the kinship matrix is defined as the variance-covariance matrix prior of the random effects, which accurately describes the genetic correlation structure. The restricted maximum likelihood method is used to fit the model, and the variance components are solved by the EM algorithm to obtain the additive genetic variance component and the residual variance component of the chelicera ratio trait. By calculating the ratio of the additive genetic variance to the total phenotypic variance, the accurate estimate of the heritability of the trait is obtained, and the best linear unbiased prediction method is used to solve the breeding value of each individual, providing a quantitative basis for subsequent selection.

[0027] On the basis of obtaining genetic parameters, a phenomics network is further constructed to realize visual analysis of multi-dimensional data. The key growth characteristics are extracted from the association rule set as network nodes, the dynamic time warping algorithm is used to calculate the phenotype distance matrix between individuals, and the similarity of different individuals in the asynchronous growth trajectory is effectively represented. At the same time, the graphicallasso algorithm is used to estimate the conditional independence relationship between the characteristics, construct a sparse Gaussian graph model, and reveal the direct correlation relationship between the characteristics. The sparse Gaussian graph model is used as the network basic topology, in which the nodes represent the phenotype characteristics and the edge weights represent the conditional correlation coefficients between the characteristics. The association strength of each characteristic with the chelicera ratio is extracted from the association rule set as the node attribute, the heritability estimate value is used as the global attribute of the network, and the individual breeding value is used as the additional attribute of the corresponding individual phenotype node, to generate an initial phenomics network rich in multi-dimensional information. Finally, the Node2Vec algorithm is used to learn the low-dimensional vector representation of the network nodes, the node sequence is generated by the biased random walk strategy, the Skip-gram model is used to learn the node embedding vector, and the t-SNE dimension reduction algorithm is used to map the high-dimensional embedding to two-dimensional space, and finally a visual macrobrachium rosenbergii phenomics network is generated. The network effectively integrates the correlation relationship between the phenotype characteristics, the genetic parameter information and the individual breeding value, and provides an intuitive multi-dimensional analysis tool for breeding decision-making.

[0028] Further, in a preferred embodiment of the present application, the growth path of Macrobrachium rosenbergii in the breeding process is analyzed by using the macrobrachium rosenbergii breeding database, abnormal events during breeding are identified and corresponding treatment schemes are associated, and a macrobrachium rosenbergii growth trend network is generated, specifically comprising: Obtaining a database of Macrobrachium rosenbergii culture, extracting time-series phenotype measurement data and corresponding environmental parameter sequences of all individual Macrobrachium rosenbergii from the database of Macrobrachium rosenbergii culture, and after preprocessing, using a nonlinear mixed effect model to fit the standard growth curve of Macrobrachium rosenbergii by taking individuals as random effects, to establish an individual growth curve model describing the normal growth pattern; Based on the established individual growth curve model, the residual sequence between the actual measured value and the model predicted value of each individual is calculated, and wavelet transform analysis is performed on the residual sequence to extract time-frequency domain features and generate a set of Macrobrachium rosenbergii individual growth trajectories; combining the environmental parameter sequence, the isolation forest algorithm is used to detect abnormal events in the growth process, and the occurrence time, event characteristics, duration and abnormal intensity index of the abnormal events are recorded to generate a set of abnormal growth event characteristics; According to the abnormal growth event characteristic set, the identified abnormal event points are time-stamped matched with the control log records in the culture control data set, if there is an artificial recorded control operation for a certain abnormal event point, the corresponding abnormal event and control measure are associated to form an abnormal event-disposal scheme association pair, otherwise it is an environmental self-fluctuation event, an abnormal event-self fluctuation association pair is generated, and it is marked as an event without disposal; Based on the obtained association pair, an abnormal-disposal mapping knowledge base is constructed in the order of abnormal event characteristics, environmental parameter characteristics and disposal scheme, including abnormal event type, environmental parameter characteristics, disposal measures taken and disposal effect; Through the set of Macrobrachium rosenbergii individual growth trajectories, the growth trajectory characteristics of each Macrobrachium rosenbergii individual are extracted to generate a number of growth trajectory characteristic sequences with time-series attributes, based on the abnormal-disposal mapping knowledge base, the abnormal event corresponding to the Macrobrachium rosenbergii individual identifier and the abnormal event occurrence timestamp are merged into the corresponding growth trajectory characteristic sequence to generate a growth trend sequence representing each Macrobrachium rosenbergii individual in the culture process; The cosine similarity and Mahalanobis distance between each growth trend sequence are calculated respectively, and weighted average and normalization processing are performed to generate a sequence merging index, which is compared with a preset threshold, if it is greater than the preset threshold, the difference between the two growth trend sequences is calculated to identify a difference sequence segment, a new sequence branch is generated in the difference sequence segment for sequence merging, and through repeated iteration and merging steps, a Macrobrachium rosenbergii growth trend network is finally generated.

[0029] It needs to be explained that first, the time series of individual phenotypic measurements (such as body length, cheliped length, body weight, etc.) and their corresponding environmental parameter sequences (such as water temperature, dissolved oxygen, pH value, etc.) were extracted from the constructed database of Macrobrachium rosenbergii culture. After preprocessing (including missing value filling, outlier removal and standardization), a nonlinear mixed effects model was used for modeling. This model takes individuals as random effect terms, which can capture the growth variation between individuals, and fits the standard growth curve of Macrobrachium rosenbergii, thereby establishing an individual growth curve model that describes the normal growth pattern, providing a benchmark reference for subsequent analysis. Based on this model, the residual sequence between the actual measured value and the model predicted value of each individual was calculated, and these residuals reflect the extent to which individual growth deviates from the normal pattern; then wavelet transform analysis was performed on the residual sequence to extract time-frequency domain features (such as energy distribution, frequency component), thereby generating a set of Macrobrachium rosenbergii individual growth trajectories, which quantifies the dynamic growth characteristics of each individual. At the same time, combined with the environmental parameter sequence, the isolation forest algorithm was used to detect abnormal events during growth, identify abnormal time segments that deviate significantly from the environmental pattern, and record the occurrence time, event characteristics (such as sudden temperature change, dissolved oxygen anomaly), duration and abnormal intensity index of abnormal events, forming a set of abnormal growth event characteristics. Next, these identified abnormal event points were time-stamped matched with the control log records in the culture control data set: if there is a manual recorded control operation (such as oxygenation, water change, drug use) for a certain abnormal event point, then the corresponding abnormal event and control measures are associated to form an "abnormal event-treatment scheme" association pair; otherwise, it is marked as an environmental self-fluctuation event, generating an "abnormal event-self fluctuation" association pair, and labeled as a non-disposal event. Based on these association pairs, an abnormal-disposal mapping knowledge base was constructed with the association order of abnormal event characteristics, environmental parameter characteristics and disposal scheme, which contains abnormal event types, environmental parameter characteristics, disposal measures taken and disposal effects (such as recovery time, growth response), providing historical experience support for subsequent decision-making.

[0030] Further, growth trajectory features (such as growth rate, acceleration, fluctuation pattern) of each individual Macrobrachium rosenbergii are extracted from the growth trajectory set of the individual Macrobrachium rosenbergii, to generate a plurality of growth trajectory feature sequences with time sequence attributes; then, based on the abnormality-disposal mapping knowledge base, the individual Macrobrachium rosenbergii identifier and the occurrence timestamp of the abnormal event are merged into the corresponding growth trajectory feature sequence, so as to generate a sequence representing the complete growth trend of each individual Macrobrachium rosenbergii in the cultivation process, which integrates multi-dimensional information of normal growth, abnormal events and disposal responses. Finally, the cosine similarity (measuring sequence shape similarity) and Mahalanobis distance (measuring sequence statistical distribution difference) between each growth trend sequence are calculated respectively, and a weighted average and normalization processing is performed to generate a sequence merging index; the index is compared with a preset threshold value, if greater than the threshold value, difference calculation is performed on the two growth trend sequences to identify a difference sequence segment (such as an event response segment), and a new sequence branch is generated in the difference sequence segment for sequence merging; by repeating the merging step, a Macrobrachium rosenbergii growth trend network is finally generated, which visually displays the similarity of different individuals or groups in the growth process, the influence path of abnormal events and the effectiveness of disposal measures, and provides data-driven insights for optimizing the cultivation strategy.

[0031] Further, in a preferred embodiment of the present application, the early phenotype data of the to-be-selected Macrobrachium rosenbergii population is collected, the selection decision is made through the Macrobrachium rosenbergii chelae ratio prediction model, and the breeding path recommendation is made for the selected individual by using the Macrobrachium rosenbergii phenotype knowledge graph, which specifically includes: When the to-be-selected Macrobrachium rosenbergii population is cultivated, the morphological phenotype data of each individual is obtained through the deployed automatic image acquisition system, and the cultivation environment parameters are synchronously monitored and collected to obtain cultivation environment monitoring data; The collected early phenotype data and cultivation environment monitoring data are preprocessed and standardized to form an early monitoring feature vector of the to-be-selected individual; the early monitoring feature vector is input into the trained Macrobrachium rosenbergii chelae ratio prediction model, and the prediction value of the chelae ratio of the individual after adulthood and the corresponding prediction confidence interval are output through the multi-layer nonlinear transformation and feature extraction inside the model; Based on the chelae ratio prediction results of all to-be-selected individuals, the individuals are sorted according to the chelae ratio prediction value, the prediction confidence interval and the preset selection intensity threshold value are combined, and the Macrobrachium rosenbergii individuals with chelae ratio meeting the preset expectation are selected as candidate parents to generate a Macrobrachium rosenbergii selection recommendation table; According to the macrobrachium recommended breeding table, combined with the macrobrachium phenotype knowledge graph for breeding path analysis, taking the early monitoring feature vector and pedigree information of the selected individual as the query condition, the historical individual nodes with similar phenotype characteristics and genetic background are retrieved in the knowledge graph, the best environmental parameter range experienced by the historical individual nodes in the complete breeding cycle, the successful control measure sequence and the final growth performance index are obtained, and the similar breeding case is generated; Based on the retrieved similar breeding case, the case reasoning technology is used to generate the individualized breeding path recommendation scheme of the target recommended breeding individual, including the environmental parameter control target, feeding strategy, density management requirement and expected abnormal events and countermeasures in different growth stages; The macrobrachium breeding recommendation table is associated with the individualized breeding path recommendation scheme to generate the final macrobrachium breeding decision report for pushing.

[0032] It should be noted that when breeding the selected breeding population, first, the key morphological phenotype data of each individual is obtained regularly by the deployed automatic image acquisition system (usually including high-definition industrial cameras and image processing units), including body length, chelae length, body weight and other indicators, and at the same time, the Internet of Things sensor network is used to monitor and collect the breeding environment parameters (such as water temperature, pH value, dissolved oxygen, ammonia nitrogen concentration, etc.) simultaneously, and obtain the breeding environment monitoring data corresponding to the phenotype data in space and time. The early multi-source heterogeneous data collected is preprocessed and standardized, including data cleaning, outlier removal, missing value filling and feature scaling, to form the standardized early monitoring feature vector of the selected breeding individual, providing high-quality input for subsequent prediction analysis. The early monitoring feature vector is input into the pre-trained macrobrachium chelae ratio prediction model, and through the multi-layer nonlinear transformation and feature extraction mechanism inside the model, the model outputs the predicted value of the individual's adult chelae ratio and its corresponding prediction confidence interval, which quantifies the uncertainty degree of the prediction result. Based on the prediction results of all selected breeding individuals, the chelae ratio prediction value is sorted, and combined with the prediction confidence interval and the preset selection intensity threshold (such as the selection proportion of the top 10%), the excellent individual with chelae ratio meeting the preset expected value is selected as the candidate parent, thereby generating the macrobrachium breeding recommendation table containing individual number, prediction value and priority.

[0033] Subsequently, according to the selection recommendation table, combined with the macrobrachium rosenbergii phenotype knowledge graph, the early monitoring feature vector and pedigree information (such as family origin, parent traits) of the selected individual are analyzed in depth, the historical individual nodes with similar phenotype characteristics and genetic background are retrieved in the knowledge graph by taking the early monitoring feature vector and pedigree information (such as family origin, parent traits) of the selected individual as the query condition, the best environmental parameter range, the successful control measure sequence and the final growth performance index experienced by these historical individuals in the complete cultivation cycle are obtained, and a similar cultivation case set with reference value is generated. Based on the retrieved similar cultivation cases, a personalized cultivation path recommendation scheme for the target recommended selected individual is generated by using case reasoning technology, which specifies the environmental parameter control target (such as suitable temperature range, dissolved oxygen threshold), feeding strategy (bait type, feeding frequency and amount), density management requirement (cultivation density adjustment scheme) and expected abnormal events (such as stress response, disease risk) and corresponding countermeasures in different growth stages in detail, forming a complete and accurate cultivation guidance scheme. Finally, the macrobrachium rosenbergii selection recommendation table and the personalized cultivation path recommendation scheme are associated and integrated to generate a structured final macrobrachium rosenbergii selection decision report, which is delivered to the cultivation manager through a visual interface or an automatic push system, providing data-driven decision support for actual selection operation.

[0034] Figure 2 A second method flowchart of a macrobrachium rosenbergii chela ratio prediction and selection method based on phenomics analysis is provided for an embodiment of the present application; As shown in Figure 2 The present application provides a second method flowchart of a macrobrachium rosenbergii chela ratio prediction and selection method based on phenomics analysis, which comprises: S202, obtaining a macrobrachium rosenbergii phenotype network and a macrobrachium rosenbergii growth trend network, extracting all phenotype feature nodes and node attribute information from the macrobrachium rosenbergii phenotype network, including the correlation strength of growth characteristics and chela ratio, genetic force estimate value and individual breeding value, and extracting abnormal event nodes, disposal scheme nodes and their association from the macrobrachium rosenbergii growth trend network; S204, based on the extracted phenotype feature nodes, taking the phenotype change time sequence of the macrobrachium rosenbergii individual growth process as the node connection order to generate a phenotype node sequence main shaft, which represents all phenotype change patterns of the macrobrachium rosenbergii individual in the growth process; connecting the abnormal event nodes and the disposal scheme nodes based on the association relationship between the abnormal event nodes and the disposal scheme nodes as the connection basis to generate a node sequence sub-shaft; S206, merging the node sequence sub-shaft with the node sequence main shaft according to the trigger event stamp of the abnormal event as the index to obtain a plurality of tree-shaped node sequences, which represent the phenotype change patterns and growth trend of the macrobrachium rosenbergii individual in the cultivation process, and finally based on all tree-shaped node sequences to constitute a macrobrachium rosenbergii phenotype knowledge graph; S208, learning low-dimensional vector representation of entities and relationships of the macrobrachium Rosenbergii phenotype knowledge graph by a TransE knowledge identification learning algorithm, building a macrobrachium Rosenbergii chelae ratio prediction model based on a graph neural network as a basic framework, training the model by using the training data set, optimizing the hyperparameters by cross-validation and grid search optimization, and finally obtaining a macrobrachium Rosenbergii chelae ratio prediction model meeting the expectation.

[0035] It should be noted that in the process of constructing the macrobrachium Rosenbergii phenotype knowledge graph, first, the macrobrachium Rosenbergii phenomics network and the macrobrachium Rosenbergii growth trend network are integrated. All phenotype feature nodes and their attribute information are extracted from the phenomics network, including the correlation strength between each growth characteristic and the chelae ratio, the genetic force estimate value, and the individual breed value and other key parameters; meanwhile, the abnormal event nodes, the disposal scheme nodes and the correlation between them are extracted from the growth trend network. These nodes and relationships constitute the basic elements of the knowledge graph. Based on the extracted phenotype feature nodes, the nodes are connected in time sequence to form a phenotype node sequence main shaft according to the time sequence relationship of the phenotype changes of the macrobrachium Rosenbergii individual growth process. At the same time, according to the causal relationship between the abnormal event nodes and the disposal scheme nodes, a sequence sub-axis connecting these nodes is generated to form an event-disposal correlation network. The network structures in the above two axes are fused, the node sequence sub-axis and the phenotype node sequence main shaft are spatiotemporally associated with each other with the timestamp of the abnormal event as the index point, and a tree-shaped node sequence with a branch structure is generated. This tree structure can intuitively show the corresponding relationship between the phenotype changes and the abnormal events and disposal measures in the individual growth process, and finally by integrating the tree-shaped node sequences of all individuals, a complete macrobrachium Rosenbergii phenotype knowledge graph is constructed. After the construction of the knowledge graph, knowledge representation learning algorithms such as TransE are used to learn low-dimensional vector representation of entities and relationships in the graph, and discrete graph information is converted into continuous vector space representation. Based on the learned vector representation, a training data set is constructed, and a chelae ratio prediction model is built based on a graph neural network as a basic framework. The model updates the node representation by aggregating the neighborhood information, which can effectively capture the complex correlation between phenotype characteristics, abnormal events and chelae ratio. Finally, the model hyperparameters are optimized by cross-validation and grid search methods, and an excellent chelae ratio prediction model is obtained, which provides reliable technical support for the precise breeding of macrobrachium Rosenbergii.

[0036] The recommended breeding of the macrobrachium Rosenbergii is continuously monitored during the breeding process, and whether the macrobrachium Rosenbergii growth trend deviates from the preset breeding expectation is analyzed through the macrobrachium Rosenbergii growth trend data, a correction scheme is developed and a warning prompt is given, which specifically includes: Figure 3A third method flowchart of a Macrobrachium rosenbergii chela ratio prediction and breeding method based on phenomics analysis is provided for an embodiment of the present application, as shown in the drawings. As Figure 3 shown, the present application provides a third method flowchart of a Macrobrachium rosenbergii chela ratio prediction and breeding method based on phenomics analysis, comprising: S302, real-time acquisition of multi-modal monitoring data of breeding individuals in the breeding process, including time-series environmental parameters collected by a sensor array and individual phenotype measurement data periodically acquired by an image acquisition system; S304, comparing the multi-modal monitoring data with the preset breeding path expected value recommended for the individual in the Macrobrachium rosenbergii phenotype knowledge graph, calculating the similarity distance between the actual growth sequence and the expected growth sequence using a dynamic time warping algorithm, and setting a deviation threshold based on a statistical process control method; S306, when it is detected that the actual growth trajectory or environmental parameter continuously deviates from the expected path and exceeds the deviation threshold, triggering an early warning mechanism; querying the abnormal-disposal mapping knowledge base in the Macrobrachium rosenbergii growth trend network according to the deviation characteristics, matching the historical similar deviation mode and the corresponding effective disposal scheme; at the same time, retrieving the genetic background information and phenotype characteristic association rules of the individual in the Macrobrachium rosenbergii phenotype knowledge graph, evaluating the potential influence of the specific disposal measures on the genetic strain; S308, comprehensively considering the effect of historical disposal scheme and individual genetic characteristics, generating an individualized correction scheme for the current deviation condition, the correction scheme including environmental parameter adjustment strategy, feeding management measure adjustment suggestion and expected recovery trajectory; pushing the correction scheme to the breeding management system and synchronously triggering the early warning prompt, guiding manual or automatic execution of control operation; S310, after executing the correction scheme, continuing to monitor the growth response and environmental parameter changes of the individual, comparing the actual recovery trajectory with the expected recovery trajectory, and evaluating the implementation effect of the correction scheme; taking the deviation event, the correction measures taken and the final effect as new knowledge samples, feeding back and updating to the Macrobrachium rosenbergii growth trend network and the Macrobrachium rosenbergii phenotype knowledge graph, realizing continuous optimization and learning of the knowledge base, forming a closed-loop intelligent control system of monitoring-early warning-decision-feedback.

[0037] It should be noted that in the intelligent breeding process of Macrobrachium rosenbergii, there will still be deviations from the recommended breeding path, which need to be corrected to avoid the final result of breeding errors. Through the deployment of sensor arrays and image acquisition systems, real-time acquisition of multi-modal monitoring data of breeding individuals is realized, including time series environmental parameters such as water temperature, dissolved oxygen, pH value, and individual phenotype measurement data such as body length and chelae length. Continuous comparison with the expected value of the preset individualized breeding path, the dynamic time warping algorithm is used to calculate the similarity distance between the actual growth sequence and the expected growth sequence, to accurately evaluate the degree of coincidence of the growth trajectory, and based on the statistical process control method, the deviation threshold is set, and the quantitative standard of growth monitoring is established. When the actual growth trajectory or environmental parameters continuously deviate from the expected path and exceed the deviation threshold, the early warning mechanism will be automatically triggered. Then according to the specific characteristics of the deviation, the abnormal-disposal mapping knowledge base in the Macrobrachium rosenbergii growth trend network is intelligently queried, and cases with similar deviation patterns and their corresponding effective disposal schemes in history are matched. At the same time, the genetic background information and phenotype characteristic association rules of the individual in the Macrobrachium rosenbergii phenotype knowledge graph are retrieved, and the potential influence and applicability of the specific disposal measures on the genetic strain are comprehensively analyzed to ensure the pertinence of the recommended scheme. On the basis of the effect of historical disposal schemes and individual genetic characteristics, individualized correction schemes are generated for the current deviation condition. The correction scheme includes adjustment strategies for environmental parameters, specific adjustment suggestions for feeding management measures, and scientific prediction of the expected recovery trajectory. The correction scheme is pushed to the breeding management system in time, and the early warning prompt is triggered at the same time, guiding the staff to perform manual control or directly triggering the automatic equipment to perform control operation. After the implementation of the correction scheme, the growth response of the individual and the change of the environmental parameters are closely monitored, the actual recovery trajectory is compared with the expected recovery trajectory, and the implementation effect of the correction scheme is evaluated. Finally, the characteristics of this deviation event, the correction measures taken and the final effect are fed back and updated to the Macrobrachium rosenbergii growth trend network and the Macrobrachium rosenbergii phenotype knowledge graph, so that the knowledge base can continuously accumulate practical experience, realize self-improvement and continuous optimization of the system, and form a complete monitoring-early warning-decision-feedback closed-loop intelligent control system, thereby improving the precision and intelligent level of Macrobrachium rosenbergii breeding.

[0038] Another aspect of the present application provides a computer readable storage medium comprising a Macrobrachium rosenbergii chelae ratio prediction and breeding method program based on phenomics analysis, wherein the Macrobrachium rosenbergii chelae ratio prediction and breeding method program based on phenomics analysis is executed by a processor to realize the steps of the Macrobrachium rosenbergii chelae ratio prediction and breeding method based on phenomics analysis according to any one of the above aspects.

[0039] The application provides a method for predicting and breeding Macrobrachium rosenbergii based on phenomics analysis, wherein a phenotype knowledge graph of the Macrobrachium rosenbergii is constructed, so as to analyze the correlation between the claw body ratio and the growth index in the growth process of the Macrobrachium rosenbergii, and meanwhile, corresponding culture data are combined to further reveal the growth situation of the Macrobrachium rosenbergii with a preset expected claw body ratio in the growth process, that is, the growth mode of the Macrobrachium rosenbergii is characterized based on different growth trends and situations of the Macrobrachium rosenbergii, so that the Macrobrachium rosenbergii with a certain type of growth mode has a growth curve different from that of the Macrobrachium rosenbergii with other types, and is prone to abnormal events (disease or stress performance, etc.) in some stages of the growth process, so that an individualized culture path recommendation is provided, the breeding decision is provided, and the corresponding culture survival rate is further improved through the culture path recommendation. Whether the culture process deviates from the culture path is analyzed, and a correction scheme is formulated and pushed, so that the breeding cycle of the Macrobrachium rosenbergii is shortened, and the accuracy and efficiency of breeding are improved.

[0040] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other manners. The above-described device embodiments are only schematic. For example, the division of the units is only a logical function division. There can be another division manner for the actual implementation, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the various components shown or discussed can be indirect coupling or communication connection through some interfaces, and can be electrical, mechanical or other forms.

[0041] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units; they can be located in one place, or distributed on multiple network units; and part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0042] In addition, each functional unit in each embodiment of the application can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be realized in the form of hardware or in the form of hardware plus software functional unit.

[0043] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware, the foregoing program can be stored in a computer readable storage medium, and the program executes the steps of the method embodiments when executed; and the foregoing storage medium includes a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disc or an optical disc, and various storage medium capable of storing program codes.

[0044] Alternatively, the integrated unit of the present application can be stored in a computer readable storage medium if it is realized in the form of a software function module and sold or used as an independent product. Based on such understanding, the technical solutions of the embodiments of the present application can be embodied in the form of a software product, and the computer software product is stored in a storage medium, includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the embodiments of the present application. The foregoing storage medium includes a mobile storage device, a ROM, a RAM, a magnetic disc or an optical disc, and various storage medium capable of storing program codes.

[0045] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for predicting and breeding the claw-to-body ratio of Macrobrachium rosenbergii based on phenomics analysis, characterized in that, include: Several giant freshwater prawns were selected for multi-generational culture. During the culture process, pedigree data of each individual was recorded, and phenotypic data and culture management data of the prawns at different culture stages were collected to construct a giant freshwater prawn culture database. Phenotypic characteristics of prawns at different farming stages were extracted from the prawn farming database, the relationship between individual growth indicators and chelicerate ratio was analyzed, heritability was assessed, and a prawn phenomics network was constructed. The growth path of giant freshwater prawns during the farming process is analyzed using the aforementioned prawn farming database. Abnormal events during the farming period are identified and corresponding treatment plans are associated with them, generating a prawn growth status network. A phenotypic knowledge graph of the prawn was constructed by combining the prawn phenomics network and the prawn growth status network, and a prawn chelicerae ratio prediction model was constructed using the prawn phenotypic knowledge graph. Early phenotypic data of the selected giant freshwater prawn population were collected. Breeding decisions were made using the giant freshwater prawn claw-to-body ratio prediction model. The giant freshwater prawn phenotypic knowledge graph was used to recommend farming paths for the selected individuals. Acquire multimodal monitoring data of selected individuals during the breeding process, analyze whether the actual growth trajectory or environmental parameters deviate from the expected path, and if deviations are found, formulate and implement corrective measures.

2. The method for predicting and breeding the claw ratio of Macrobrachium rosenbergii based on phenomics analysis according to claim 1, characterized in that, The process involves selecting several giant freshwater prawns for multi-generational culture, recording the pedigree data of each individual during the culture process, and collecting phenotypic data and culture management data of the prawns at different culture stages to construct a giant freshwater prawn culture database. Specifically, this includes: Several giant freshwater prawns with a chelicer-to-body ratio that meet the desired threshold are selected from existing culture ponds as a first-generation culture population. Each giant freshwater prawn parent and offspring is given a unique electronic identifier. A culture pedigree of giant freshwater prawns is constructed through multiple generations of culture. During the aquaculture process, an array of aquaculture monitoring sensors is deployed to periodically collect individual image data of prawns at different growth stages, environmental data of prawn farming ponds, and prawn management data to generate a raw multimodal dataset. The obtained original multimodal dataset is preprocessed, and the collected individual images of prawns are imported into an image segmentation model built with a convolutional neural network to obtain individual prawn segmentation images. The morphological phenotypic data of the individuals are calculated according to the preset pixel-actual size conversion coefficient and bound to the corresponding individual ID and collection timestamp to generate a prawn phenotypic dataset. For the environmental data and management data of prawn farming ponds, an outlier monitoring algorithm is used to clean the data, identify and remove abnormal measurement values, extract the environmental data fluctuation characteristics before and after the event point based on the timestamp of the management event recorded in the farming log, and generate a management tag for each event. The management tag includes management measures, changes in environmental parameters before and after management, and prawn status data, and finally generates a farming management dataset with event semantics. A database of giant freshwater prawn (Gastropoda rosenbergii) farming was constructed by combining the pedigree of giant freshwater prawn farming, prawn phenotypic datasets, and farming management datasets.

3. The method for predicting and breeding the claw ratio of Macrobrachium rosenbergii based on phenomics analysis according to claim 1, characterized in that, The step of extracting phenotypic characteristics of prawns at different farming stages from the prawn farming database and analyzing the relationship between individual growth indicators and cheliped body ratio specifically includes: Obtain the prawn farming database, extract complete time-series phenotypic data of all prawn individuals at various time points from the prawn farming database, generate prawn phenotypic measurement sequences, and use a sliding window of fixed time length to segment the prawn phenotypic measurement sequences to generate several prawn phenotypic measurement subsequences. For each phenotypic measurement subsequence of freshwater prawns, the first-order difference between adjacent time points is calculated as the instantaneous growth rate characteristic, and the difference between the first-order difference sequences is calculated as the second-order difference to characterize the growth acceleration. Meanwhile, cubic spline interpolation or polynomial fitting is performed on the complete growth curve of each phenotypic index, the coefficients of the fitting function are extracted as shape features describing the overall growth pattern, and the statistical features of each sequence are calculated, including mean, variance, skewness and kurtosis, and finally the phenotypic derived feature sequence of prawn is obtained. A multimodal feature set characterizing individual growth dynamics was constructed by integrating the phenotypic measurement sequence and the phenotypic derived feature sequence of the prawn. Correlation analysis was performed by combining the final measured chelicerae ratio of each prawn individual to calculate the maximum mutual information coefficient and Pearson correlation coefficient between each feature and the chelicerae ratio. Subsequently, a random forest regression model was invoked, with the final measured claw-to-body ratio as the target variable and the phenotypic measurement features and extended phenotypic features of the prawn as decision variables. The importance score of each feature was calculated through 10-fold cross-validation. The key growth features and weight coefficients that are significantly related to the claw-to-body ratio were analyzed. A Gaussian process regression model was established by selecting key growth features to fit the nonlinear mapping relationship between phenotypic features and target traits, and finally, a set of association rules between phenotypic features and claw-to-body ratio was generated.

4. The method for predicting and breeding the claw ratio of Macrobrachium rosenbergii based on phenomics analysis according to claim 1, characterized in that, The genetic assessment and construction of the prawn phenomics network specifically include: Obtain the phenotypic feature-claw body ratio association rule set, extract key growth features and weight coefficients that are significantly related to the claw body ratio from the phenotypic feature-claw body ratio association rule set to construct a fixed-effects design matrix, extract the pedigree table from the prawn farming database, and use the pedigree table to construct a kinship matrix of all prawn individuals; Based on the fixed effects design matrix and kinship matrix, the additive genetic effects of individual prawns are treated as random effects, and the kinship matrix is ​​defined as the variance-covariance matrix prior of the random effects to construct a mixed linear model based on an animal model. The restricted maximum likelihood method was used to fit the mixed linear model. The variance components were solved iteratively by the EM algorithm to obtain the additive genetic variance component and residual variance component of the chelate ratio trait. The heritability estimate was obtained by calculating the ratio of additive genetic variance to phenotypic variance. The breeding value of each individual was solved by the BLUP method. Based on the phenotypic feature-turn ratio association rule set, key growth features are extracted as network nodes. The dynamic time warping algorithm is used to calculate the phenotypic distance matrix between individuals to represent the asynchronous growth trajectory of individual prawns. The graphical lasso algorithm is used to estimate the conditional independence relationship between features and to construct a sparse Gaussian graph model. The sparse Gaussian graph model is defined as the basic topology of the phenomics network, where nodes represent phenotypic features and edge weights represent the conditional correlation coefficients between features. The association strength between each phenotypic feature and the cheliceroid ratio is extracted as the initial attribute of the node through the phenotypic feature-cheliceroid ratio association rule set. The heritability estimate is used as the global attribute of the network, and the individual breeding value is used as the additional attribute of the corresponding individual phenotypic node to generate the initial phenomics network. The Node2Vec algorithm is used to learn the low-dimensional vector representation of the nodes in the initial phenomics network. The transition probabilities between nodes are defined. Node embedding vectors are learned by generating node sequences through biased random walks. Finally, the t-SNE dimensionality reduction algorithm is used to map the high-dimensional node embeddings to a two-dimensional space to generate the prawn phenomics network.

5. The method for predicting and breeding the claw ratio of Macrobrachium rosenbergii based on phenomics analysis according to claim 1, characterized in that, The process of analyzing the growth path of giant freshwater prawns during the farming process using the prawn farming database, identifying abnormal events during the farming period and associating them with corresponding treatment plans, and generating a prawn growth status network specifically includes: A database of giant freshwater prawns was obtained. The time-series phenotypic measurement data and corresponding environmental parameter sequences of all giant freshwater prawns were extracted from the database. After preprocessing, a nonlinear mixed-effects model was used, with individuals as random effects, to fit the standard growth curve of giant freshwater prawns and establish an individual growth curve model describing the normal growth pattern. Based on the established individual growth curve model, the residual sequence between the actual measured value and the model predicted value of each individual is calculated, and wavelet transform analysis is performed on the residual sequence to extract time-frequency domain features and generate a set of individual growth trajectories of prawns; combined with the environmental parameter sequence, the isolated forest algorithm is used to detect abnormal events in the growth process, and the occurrence time, event characteristics, duration and abnormal intensity index of abnormal events are recorded to generate a set of abnormal growth event features; Based on the feature set of abnormal growth events, the identified abnormal event points are matched with the control log records in the aquaculture control dataset by timestamp. If a certain abnormal event point has a manually recorded control operation, the corresponding abnormal event is associated with the control measures to form an abnormal event-control plan association pair. Otherwise, if it is an environmental self-fluctuation event, an abnormal event-self-fluctuation association pair is generated and marked as an event that does not require control. Based on the obtained association pairs, an anomaly-handling mapping knowledge base is constructed with the association order of anomaly event characteristics, environmental parameter characteristics, and handling solutions. It includes anomaly event types, environmental parameter characteristics, handling measures taken, and the effects after handling. The growth trajectory features of each individual prawn are extracted from the prawn individual growth trajectory set to generate several growth trajectory feature sequences with time-series attributes. Based on the anomaly-handling mapping knowledge base, the prawn individual identifier corresponding to the abnormal event and the timestamp of the abnormal event are merged into the corresponding growth trajectory feature sequence to generate a sequence representing the growth status of each prawn individual in the farming process. The cosine similarity and Mahalanobis distance between each growth status sequence are calculated separately, and weighted average and normalization are performed to generate a sequence merging index. This index is then compared with a preset threshold. If the index is greater than the preset threshold, the difference between the two growth status sequences is calculated to identify the difference sequence segments. New sequence branches are generated in the difference sequence segments for sequence merging. Through repeated iterative merging steps, the prawn growth status network is finally generated.

6. The method for predicting and breeding the claw ratio of Macrobrachium rosenbergii based on phenomics analysis according to claim 1, characterized in that, The process of constructing a phenotypic knowledge graph of the giant freshwater prawn by combining the prawn phenomics network and the giant freshwater prawn growth status network, and then using the prawn phenotypic knowledge graph to construct a prawn claw-to-body ratio prediction model, specifically includes: Obtain the phenomics network and growth status network of the prawn. Extract all phenotypic feature nodes and node attribute information from the prawn phenomics network, including the correlation strength between growth characteristics and chelicerate ratio, heritability estimate and individual breeding value. Extract abnormal event nodes, treatment plan nodes and their correlation relationships from the prawn growth status network. Based on the extracted phenotypic feature nodes, the phenotypic change sequence during the growth process of individual prawns is used as the node connection order to generate the phenotypic node sequence main axis. The phenotypic node sequence main axis represents all phenotypic change patterns of individual prawns during the growth process. The correlation between abnormal event nodes and treatment plan nodes is used as the connection basis to generate node connection edges to connect abnormal event nodes and treatment plan nodes, thus generating node sequence sub-axis. Based on the node sequence sub-axis, several tree-like node sequences are obtained by merging with the node sequence main axis using the trigger event stamp of the abnormal event as the index. The tree-like node sequences represent the phenotypic change patterns and growth status of individual prawns during the farming process. Finally, a prawn phenotypic knowledge graph is constructed based on all the tree-like node sequences. The TransE knowledge identifier learning algorithm is used to learn low-dimensional vector representations of entities and relationships in the phenotypic knowledge graph of the giant freshwater prawn and construct a training dataset. A giant freshwater prawn claw-to-body ratio prediction model is built based on a graph neural network architecture, and the model is trained using the training dataset. Hyperparameters are optimized through cross-validation and grid search optimization, and finally, a giant freshwater prawn claw-to-body ratio prediction model that meets the expectations is obtained.

7. The method for predicting and breeding the claw ratio of Macrobrachium rosenbergii based on phenomics analysis according to claim 1, characterized in that, The process involves collecting early phenotypic data from the proposed Macrobrachium rosenbergii population, making breeding decisions using the Macrobrachium rosenbergii claw-to-body ratio prediction model, and recommending farming pathways for selected individuals using the Macrobrachium rosenbergii phenotypic knowledge graph. Specifically, this includes: When raising the population of giant freshwater prawns to be bred, morphological phenotypic data of each individual is acquired through an automated image acquisition system, and environmental parameters of the breeding environment are monitored and collected simultaneously to obtain breeding environment monitoring data. The collected early phenotypic data and aquaculture environment monitoring data are preprocessed and standardized to form early monitoring feature vectors of individuals to be selected for breeding. The early monitoring feature vectors are input into a pre-trained prawn claw-to-body ratio prediction model. Through multi-layer nonlinear transformation and feature extraction within the model, the predicted value of the claw-to-body ratio after adulthood and the corresponding prediction confidence interval are output. Based on the predicted chelicerae ratio of all individuals to be bred, they are sorted according to the predicted chelicerae ratio. Combined with the prediction confidence interval and the preset selection intensity threshold, individuals of prawns whose chelicerae ratio meets the preset expectation are selected as candidate parents, and a prawn breeding recommendation table is generated. Based on the recommended breeding table for freshwater prawns, and combined with the prawn phenotypic knowledge graph, the breeding path is analyzed. Using the early monitoring feature vectors and pedigree information of the individuals to be bred as query conditions, historical individual nodes with similar phenotypic characteristics and genetic backgrounds are retrieved in the knowledge graph. The optimal environmental parameter range, successful control measures sequence, and final growth performance indicators experienced by the historical individual nodes in the complete breeding cycle are obtained, and similar breeding cases are generated. Based on similar breeding cases retrieved, case reasoning technology is used to generate personalized breeding path recommendation schemes for target recommended breeding individuals, including environmental parameter control targets, feeding strategies, density management requirements, and expected abnormal events and corresponding measures for different growth stages; The prawn breeding recommendation form is linked with the personalized aquaculture path recommendation plan to generate a final prawn breeding decision report, which is then pushed out.

8. The method for predicting and breeding the claw ratio of Macrobrachium rosenbergii based on phenomics analysis according to claim 1, characterized in that, The process involves acquiring multimodal monitoring data of selected individuals during the breeding process, analyzing whether the actual growth trajectory or environmental parameters deviate from the expected path, and if deviations are found, developing and implementing corrective measures, specifically including: Real-time acquisition of multimodal monitoring data of selected individuals during the breeding process, including time-series environmental parameters collected by sensor arrays and individual phenotypic measurement data acquired periodically by image acquisition system; The multimodal monitoring data is compared with the expected value of the preset breeding path recommended for the individual in the prawn phenotypic knowledge graph. The similarity distance between the actual growth sequence and the expected growth sequence is calculated using the dynamic time warping algorithm, and the deviation threshold is set based on the statistical process control method. When the actual growth trajectory or environmental parameters are detected to continuously deviate from the expected path and exceed the deviation threshold, an early warning mechanism is triggered; based on the deviation characteristics, the abnormal-treatment mapping knowledge base in the prawn growth status network is queried to match historical similar deviation patterns and their corresponding effective treatment solutions; at the same time, the genetic background information and phenotypic feature association rules of the individual are retrieved from the prawn phenotypic knowledge graph to assess the potential impact of specific treatment measures on the genetic strain. Based on the combined effects of historical treatment plans and individual genetic characteristics, a personalized correction plan is generated for the current deviation. The correction plan includes environmental parameter adjustment strategies, suggestions for adjusting feeding and management measures, and expected recovery trajectory. The correction plan is then pushed to the aquaculture management system and an early warning is triggered simultaneously to guide manual or automatic control operations. After implementing the corrective measures, we continue to monitor the growth response of individuals and changes in environmental parameters, compare the actual recovery trajectory with the expected recovery trajectory, and evaluate the effectiveness of the corrective measures. We use the deviation event, the corrective measures taken, and the final effect as new knowledge samples, and feed them back to update the prawn growth status network and prawn phenotypic knowledge graph to achieve continuous optimization and learning of the knowledge base.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a program for predicting and breeding the claw-to-body ratio of Macrobrachium rosenbergii based on phenomics analysis. When the program for predicting and breeding the claw-to-body ratio of Macrobrachium rosenbergii based on phenomics analysis is executed by a processor, it implements the steps of the method for predicting and breeding the claw-to-body ratio of Macrobrachium rosenbergii based on phenomics analysis as described in any one of claims 1 to 8.

Images