Efficient and accurate wheat parent matching method and system based on distance between core germplasm and genome
By constructing core germplasm and genetic subpopulations, and screening for complementary genomic distances across subpopulations to optimize wheat parental combinations, the problems of poor reproducibility and high cost in traditional methods have been solved. This has enabled efficient and precise wheat parental selection, improving the scientific rigor and practicality of breeding.
Patent Information
- Application Number
- CN202511798714.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-02
- Publication Date
- 2026-03-17
AI Technical Summary
Traditional wheat parent selection methods rely on breeders' experience, resulting in poor reproducibility, high costs, and long cycles. Existing molecular marker-assisted selection lacks consideration of the whole genome genetic background, and existing technologies have computational redundancy and decision-making bias, making it impossible to achieve efficient and accurate parent selection.
A wheat parental selection method based on core germplasm and genomic distance was adopted. By constructing core germplasm, dividing genetic subpopulations, calculating genomic distance, screening parental combinations with cross-subpopulation complementarity and genomic distance optimization, optimizing genetic diversity using Core Hunter and ADMIXTURE software, and writing scripts to automatically screen the optimal combination.
It enables the preservation of high genetic diversity with extremely small sample sizes, improves the accuracy and efficiency of matching, provides a seamless transition from data analysis to field work, and enhances the scientific nature and targeted nature of breeding.
Smart Images
Figure CN121687178A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a method and system for selecting wheat parents, specifically to an efficient and precise method and system for selecting wheat parents based on core germplasm and genomic distance, belonging to the field of hybridization breeding technology. Background Technology
[0002] Traditional hybridization breeding methods rely heavily on the practical experience of breeders. Breeders subjectively select parents for hybridization by observing crop phenotypic traits, considering their lineage and historical combining ability. This method has significant limitations: First, it is highly dependent on personal experience, has poor reproducibility, and is prone to technology loss due to a talent gap; second, it lacks predictability, requiring large-scale, repeated hybridization and progeny testing to find superior genotype recombination, resulting in a large workload, long cycle, and high cost, representing a typical model of quantity over quality and low yield.
[0003] In recent years, although molecular marker-assisted selection (MAS) has been applied, it has mostly targeted single genes and lacks a comprehensive consideration of the whole genome genetic background.
[0004] Existing patents related to parental selection mostly focus on phenotypic clustering or simple genetic distance calculation, failing to effectively address the core pain point in the industry: how to efficiently and accurately screen for the optimal hybrid combination while preserving the genetic diversity of the population to the greatest extent possible.
[0005] The shortcomings of existing technologies are mainly reflected in two aspects: system structure and decision-making process. (1) Structural redundancy and low efficiency: The existing scheme lacks an effective dimensionality reduction preprocessing step and directly performs full pairing analysis on a large-scale population, resulting in huge computational redundancy, making the method cumbersome and difficult to put into practical use. (2) One-sided decision-making and poor accuracy: Existing schemes rely on single-dimensional information (such as population structure or pursuing the maximum genetic distance), which cannot simultaneously optimize the two key objectives of hybrid vigor potential and complementary superior genes, resulting in blind selection results and insufficient predictability.
[0006] In summary, existing technologies, due to structural redundancy and biased decision-making, cannot achieve efficient and accurate selection of crop parents. Summary of the Invention
[0007] To address the shortcomings of existing technologies, the present invention aims to provide a method and system for efficient and precise selection of wheat parents based on core germplasm and genomic distance.
[0008] To achieve the above objectives, the present invention adopts the following technical solution: A highly efficient and precise method for selecting wheat parents based on core germplasm and genomic distance includes the following steps: (1) Obtain whole-genome SNP genotype data of the breeding population; (2) Based on whole-genome SNP genotype data, construct a core germplasm with a smaller sample size than the original population but retaining ≥95% of the original genetic diversity; (3) Analyze the population genetic structure of the core germplasm and divide it into genetic subgroups; (4) Calculate the genomic distances between core germplasm individuals; (5) The primary condition is that the parents come from different genetic subpopulations, and the initial screening range is the range of genomic distance from the upper quartile to the maximum value. Within this range, the Top N combinations with the largest and smallest genomic distances are further screened, where N is the number of output combinations required. (6) Output a list of recommended hybridization combinations.
[0009] Preferably, in step (2), the sample size of the core germplasm is 15%-25% of the original population.
[0010] Preferably, in step (4), the genomic distance is the normalized Manhattan distance.
[0011] A highly efficient and precise selection system for wheat parents, the selection system being able to implement the aforementioned selection method, and the selection system comprising the following functional modules: Data acquisition and storage module: This module is responsible for receiving, verifying and storing raw whole-genome SNP genotype data transmitted from external detection platforms, and converting it into a standardized data matrix that the system can process; Core germplasm optimization construction module: This module is responsible for receiving genotype data from the entire population, running the built-in optimization algorithm, and selecting the optimal core germplasm subset from the original population with the goals of genetic diversity coverage ≥95% and minimizing the sample size; Multidimensional genetic analysis module: This module performs two analysis tasks in parallel. One analysis task is the population genetic structure analysis performed by the population structure resolution unit on the core germplasm, which divides the genetic subpopulations and assigns a subpopulation label to each sample. The other analysis task is the pairwise genomic distance calculation performed by the genomic distance calculation unit on the core germplasm, which generates a distance matrix. Intelligent parental selection decision module: This module is responsible for executing the core decision-making logic of first qualitative and then quantitative analysis. Based on the preset rules of first screening cross-subpopulation combinations and then optimizing genomic distance, it sorts all possible combinations according to genomic distance and selects the optimal parental combination. Hybridization scheme generation and output module: This module is responsible for converting machine-readable screening results into schemes that can be directly understood and executed by those skilled in the art, generating a clear and structured list of recommended hybridization combinations; The above functional modules work together through data flow and control logic, specifically: The data acquisition and storage module is the starting point of the system. Its output is directly input into the core germplasm optimization and construction module. After the core germplasm optimization and construction module completes its processing, it outputs the generated core germplasm genotype data to the multidimensional genetic analysis module. After the multidimensional genetic analysis module completes its processing, it outputs the population structure division results and genomic distance matrix to the intelligent parental selection and decision-making module. After the intelligent parental selection and decision-making module completes its processing, it outputs the final list of optimal parental combinations to the hybridization scheme generation and output module. The hybridization scheme generation and output module is the end point of the system.
[0012] The advantages of this invention are: (1) Regarding the construction of core germplasm: Traditional methods either involve random sampling or rely solely on phenotypic clustering, which lacks representativeness. This invention utilizes Core Hunter 3.0, a professional tool, to optimize the process using allele diversity and corrected Roger distance (EN_MR) as quantitative targets. This ensures that over 95% (actual value: 99.8%) of genetic diversity is efficiently represented with only 20% of the sample size, scientifically resolving the contradiction between resource preservation and utilization efficiency.
[0013] (2) Parental evaluation dimensions: Existing technologies either only focus on population structure (e.g., avoiding inbreeding) or only consider genetic distance (e.g., pursuing maximum distance). This invention innovatively couples the population structure results of ADMIXTURE analysis with the quantitative index of Manhattan distance through an algorithm model, realizing dual screening of subpopulation complementarity and optimal distance, making the decision-making logic more scientific and the prediction accuracy higher.
[0014] (3) Technical implementation: Traditional breeding decisions rely on manual review of pedigrees and phenotypic records, which is inefficient. This invention transforms complex genetic analysis into an automated and repeatable screening process by writing scripts, and finally outputs a checklist that can directly guide field work, achieving seamless integration from data analysis to breeding practice. It is a concrete realization of the precision breeding concept. Attached Figure Description
[0015] Figure 1 This is a PCA clustering diagram of 370 original germplasm accessions; Figure 2 This is a schematic diagram showing the population genetic structure of 74 core germplasm accessions; Figure 3 This is a normalized Manhattan distance distribution histogram of 74 core germplasm accessions. Detailed Implementation
[0016] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.
[0017] I. Methods for Selecting and Matching Wheat Parent Lines The wheat parental selection method provided by this invention, based on core germplasm and genomic distance, achieves efficient and precise transformation from genomic data to breeding decisions through the systematic integration of multiple key steps. The selection method specifically includes the following steps: 1. Obtain whole-genome SNP genotype data of the breeding population. High-throughput sequencing or SNP microarray analysis was performed on the breeding population to obtain high-density SNP markers covering the entire genome of the breeding population. The raw data underwent quality control, filtering, and genotyping to finally obtain a standardized 0 / 1 genotype matrix containing all samples (0 represents the reference allele and 1 represents the variant allele).
[0018] 2. Constructing core germplasm The genotype matrix was imported into Core Hunter 3.0 software with the goal of improving allele diversity. The optimization target was set to retain ≥95% of the original genetic diversity, and the core germplasm sample size was set to 15-25% of the original population. The software's built-in optimization algorithm was used to perform multiple iterations to select the optimal sample subset that could meet the above conditions, which is the core germplasm.
[0019] This step redefines core germplasm technology from its role in the permanent preservation of germplasm resources to a preprocessing and dimensionality reduction tool serving efficient breeding decisions. It captures the maximum genetic information (≥95% diversity) with minimal analysis cost (only 15-25% of samples), solving the problem of computational redundancy in whole-population analysis and providing an efficient data foundation for subsequent precise analysis.
[0020] 3. Analyze the population genetic structure of core germplasm. The genotype data of the core germplasm were input into the ADMIXTURE software, and the optimal number of ancestral populations (K value) was determined through cross-validation. The core germplasm was then divided into several subpopulations with different genetic backgrounds.
[0021] This step aims to clarify the kinship and genetic differentiation among germplasms.
[0022] 4. Calculate the genomic distance between core germplasm individuals Based on the SNP matrix of the core germplasm, the genomic distance between all pairs of samples within the core germplasm is calculated.
[0023] 5. Select the optimal parent combination Write a script to automatically iterate through all possible combinations of core germplasm and select the optimal parental combination. The specific selection criteria are as follows: (1) Cross-subpopulation complementarity (primary condition): The parents must come from different subpopulations divided by ADMIXTURE (the purpose is to maximize genetic heterogeneity and heterosis potential). This screening condition lays a solid foundation for the generation of strong heterosis.
[0024] (2) Distance screening: Under the premise of satisfying cross-subpopulation complementarity, further screening is conducted to select combinations whose genomic distance is within a set range—from the upper quartile to the maximum value. This screening condition helps the selected parents to achieve better complementarity of superior genes while possessing heterosis potential.
[0025] (3) Among the combinations that are complementary across subgroups and whose genomic distance is within the above-defined range, select the Top N combinations with the largest (maximum complementarity) and smallest (nearest complementarity) genomic distances respectively (N is the number of output combinations required, which can be adjusted as needed, such as Top 5, Top 10, etc.). This selection condition helps to improve the degree of genomic complementarity, thereby increasing the probability of offspring exhibiting superparental traits.
[0026] This step overcomes the limitations of screening based on a single indicator (such as pursuing only the maximum genetic distance). Through multi-dimensional collaborative analysis that combines qualitative and quantitative methods, it ensures that the final recommended hybridization combination is the optimal solution of the system, rather than the result of local optima or blind selection, thus significantly improving the accuracy of the selection.
[0027] Furthermore, this step, through the development of R language scripts, integrates the core germplasm construction, population structure analysis, genomic distance calculation, and optimal parent selection processes into an automated workflow, realizing the direct transformation from complex analysis to field application, and significantly improving the practicality and scalability of the method.
[0028] The following section uses 370 wheat germplasm resources as examples to provide a detailed description of the wheat parent selection method provided by this invention.
[0029] 1. Obtain whole-genome SNP genotype data of the breeding population We obtained 370 domestically approved wheat varieties and high-generation lines (see Tables 1-1, 1-2, 1-3, 1-4, 1-5, 1-6 and 1-7 for germplasm resources) as the initial breeding population.
[0030] Genotypic analysis of the population was performed using a wheat 55K SNP chip to obtain high-density SNP markers covering the entire genome.
[0031] Quality control was performed on the raw SNP data to remove loci with low detection rates and abnormal heterozygosity, ultimately yielding 40,815 high-quality SNP loci. To further reduce the interference of linkage disequilibrium (LD) on the analysis, LD filtering was performed using PLINK software (parameters: --indep-pairwise 50 10 0.2), obtaining 1,706 high-quality, non-redundant SNP loci. This resulted in a final, standardized 0 / 1 genotype matrix containing all samples for subsequent analysis, where 0 represents the reference allele and 1 represents the variant allele.
[0032] Table 1-1 Codes and Names of 370 Wheat Germplasm Resources (Part 1)
[0033] Table 1-2 Codes and Names of 370 Wheat Germplasm Resources (Part II)
[0034] Table 1-3 Codes and Names of 370 Wheat Germplasm Resources (Part III)
[0035] Table 1-4 Codes and Names of 370 Wheat Germplasm Resources (Part Four)
[0036] Table 1-5 Codes and Names of 370 Wheat Germplasm Resources (Part 5)
[0037] Table 1-6 Codes and Names of 370 Wheat Germplasm Resources (VI)
[0038] Table 1-7 Codes and Names of 370 Wheat Germplasm Resources (VII)
[0039] 2. Constructing core germplasm The genotype matrix of the aforementioned 1706 SNPs was imported into Core Hunter 3.0 software. The optimization objective was to improve allele diversity (i.e., allele retention rate) (MR being the distance parameter). The optimization goal was set to retain ≥95% of the original genetic diversity, and the sample size of the core germplasm was set to 20% of the original population (i.e., 74 accessions). The software's built-in optimization algorithm (greedy algorithm) was used for multiple iterative searches to select the optimal sample subset that met the above conditions. Finally, core germplasm containing 74 materials was efficiently selected from the large original population. The list of the 74 core germplasm accessions is shown in Tables 2-1 and 2-2.
[0040] Table 2-1 List of 74 core germplasm accessions (Part 1)
[0041] Table 2-2 List of 74 core germplasm accessions (Part II)
[0042] Evaluation revealed that the allele retention rate of the core germplasm was as high as 99.83%, far exceeding the 95% threshold set in this invention, fully demonstrating the superior effect of the method of this invention in preserving almost complete original genetic diversity even with extremely small sample sizes. PCA clustering of 370 original germplasms was performed and the core germplasm was labeled, showing the distribution of the core germplasm in each group (…). Figure 1 ).
[0043] 3. Analyze the population genetic structure of core germplasm. Genotypic data of the core germplasms were input into ADMIXTURE software for population genetic structure analysis. Cross-validation was used to determine the optimal ancestral population size (K=4). Based on this, the 74 core germplasms were divided into four subpopulations (Pop1, Pop2, Pop3, Pop4) with different genetic backgrounds, containing 5, 7, 12, and 50 accessions respectively. Figure 2 ).
[0044] This step provides a crucial classification basis for achieving cross-subgroup matching.
[0045] 4. Calculate the genomic distance between core germplasm individuals Based on the SNP matrix (1706 SNP loci) of 74 core germplasms, the Manhattan distance between all pairs of samples within the core germplasms was calculated, and the calculation results were normalized to eliminate the bias caused by the number of markers and ensure the comparability of genetic distance values.
[0046] The normalized Manhattan distance distribution histogram of 74 core germplasm accessions is shown below. Figure 3 .
[0047] Depend on Figure 3 It can be seen that the calculated genomic distance covers a wide range. The normalized Manhattan distance of the farthest pair of materials (A245 vs A259) is 1786, and the normalized Manhattan distance of the closest pair of materials (A157 vs A220) is 1060.
[0048] The above results indicate that significant genetic differences exist within the core germplasm. This provides ample room for further optimization and screening.
[0049] 5. Select the optimal parent combination Write an R (or Python) script to automatically iterate through all possible combinations of core germplasm and select the optimal parental combination. The specific selection criteria are as follows: (1) Cross-subpopulation complementarity: The parents must come from different subpopulations divided by ADMIXTURE; (2) Distance screening: Under the premise of satisfying cross-subgroup complementarity, further screening is carried out for combinations whose genomic distance is within a set range - from the upper quartile to the maximum value; (3) Among the combinations that are complementary across subgroups and whose genomic distance is within the above-defined range, the top 5 combinations with the largest and smallest normalized Manhattan distances are selected respectively.
[0050] Finally, the R script automatically outputs two lists of recommended hybridization combinations, as detailed in Tables 3 and 4.
[0051] Table 3. Optimal 5 pairs of maximum complementary hybridization combinations and recommended priority order
[0052] Table 4. Optimal 5 pairs of most recently complementary hybridization combinations and recommended priority order
[0053] The average normalized Manhattan distance of the five pairs of combinations in Table 3 is 1705, which is recommended for creating a wide range of genetic variation and heterosis.
[0054] The average normalized Manhattan distance of the five pairs of combinations in Table 4 is 1105.4, which is recommended for maintaining specific traits and genetic stability.
[0055] These two lists of recommended hybridization combinations can be used directly to guide the design of hybridization combinations in field hybridization nurseries, allowing for the selection of appropriate parental combination strategies based on breeding objectives, thus greatly improving the targeting and efficiency of hybridization work.
[0056] In summary, the wheat parental selection method provided by this invention employs a systematic strategy of dimensionality reduction, structural analysis, and optimization. First, it constructs core germplasm that maintains ≥95% genetic diversity even with a very small sample size (e.g., 20%) using high-throughput SNP data, achieving efficient and intensive use of breeding resources. Then, it integrates population structure division and genomic distance calculation to establish a two-layer screening strategy of qualitative followed by quantitative analysis, prioritizing cross-subpopulation combinations and selecting parental pairs with optimal genetic complementarity. This method fundamentally changes the traditional experience-dependent, low-yield breeding model, providing grassroots breeding units with a direct path from massive genomic data to simple and operable field hybridization designs. It realizes a paradigm shift from manual trial and error to data-driven approaches, significantly improving the scientific rigor, relevance, and efficiency of cross-parental breeding.
[0057] II. Wheat Parental Selection System The wheat parent selection system provided by this invention, which implements the aforementioned wheat parent selection method, is data-driven and achieves automated decision-making by integrating multiple genetic analysis tools into a logical framework with a clear breeding orientation. The selection system specifically includes the following logically interconnected functional modules: 1. Data Acquisition and Storage Module The data acquisition and storage module is responsible for receiving, verifying and storing raw whole-genome SNP genotype data (usually VCF or similar format files) transmitted from external detection platforms (such as high-throughput sequencers or SNP chip scanners), and converting them into a standardized data matrix (such as a 0 / 1 matrix) that the system can process.
[0058] This module can be implemented based on common data interfaces and database technologies, such as server hard disk arrays, MySQL / PostgreSQL databases, etc.
[0059] 2. Core Germplasm Optimization Construction Module The core germplasm optimization module is key to the dimensionality reduction achieved in this invention. Its function is to receive genotype data from the entire population, run a built-in optimization algorithm, and select the optimal core germplasm subset from the original population with the objectives of achieving genetic diversity coverage ≥95% (a quantitative guarantee of high representativeness, ensuring controllable information loss after dimensionality reduction) and minimizing the sample size (the specific objective of minimization, directly determining the efficiency of dimensionality reduction). This represents a novel application of multi-objective optimization problems to dimensionality reduction in breeding resources.
[0060] The core algorithm of this module can be implemented using the functionality of existing software libraries. For example, its optimization logic can be encapsulated based on the core algorithmic ideas of tools such as CoreHunter 3.
[0061] 3. Multidimensional genetic analysis module The multidimensional genetic analysis module performs two analysis tasks in parallel: (1) Population structure analysis unit: Perform population genetic structure analysis on the core germplasm (e.g., using the core principle of the ADMIXTURE algorithm), divide the genetic subpopulations, and assign subpopulation labels to each sample; (2) Genome distance calculation unit: calculates the genomic distance between all samples in the core germplasm and generates a distance matrix.
[0062] The analysis function of this module can integrate the computational kernel of existing mature algorithms, such as the mathematical model of the ADMIXTURE software for structural partitioning.
[0063] 4. Intelligent parent selection and matching decision module The intelligent parent selection and mating decision module is the brain of this invention, responsible for executing the core decision-making logic (qualitative first, then quantitative). Based on preset rules (first screening cross-subpopulation combinations (which is the fundamental guarantee for generating hybrid vigor), and second optimizing genomic distance (which is an optimization method to achieve precise complementarity), this specific and irreversible decision-making order is the key to achieving high-precision prediction), it sorts all possible combinations according to genomic distance and selects the optimal parent combination.
[0064] This module is the core innovation of this invention. Its decision-making logic and model are entirely new and are usually implemented by custom scripts (such as R language / Python scripts).
[0065] 5. Hybridization scheme generation and output module The hybridization scheme generation and output module converts machine-readable screening results into schemes that can be directly understood and executed by those skilled in the art, generating a clear and structured list of recommended hybridization combinations (such as PDF and Excel formats).
[0066] This module can be implemented using common report generation tools or libraries, such as Python's Pandas library or R's knitr package.
[0067] The above functional modules work together through data flow and control logic, specifically: The data acquisition and storage module is the system's starting point, providing a single data source for all downstream functional modules. Its output (standardized genotype matrix) is directly input into the core germplasm optimization construction module. The core germplasm optimization construction module receives data from the data acquisition and storage module, processes it, and outputs the generated core germplasm genotype data to the multi-dimensional genetic analysis module. The multi-dimensional genetic analysis module receives data from the core germplasm optimization construction module, processes it, and simultaneously outputs both population structure division results and genomic distance matrix results to the intelligent parental selection decision module. The intelligent parental selection decision module receives both types of data from the multi-dimensional genetic analysis module, processes them using a built-in screening model, and outputs the final list of optimal parental combinations to the hybridization scheme generation and output module. The hybridization scheme generation and output module is the system's endpoint. It receives the list data from the intelligent parental selection decision module, formats it, and provides it to the user through a user interface (UI) or API. Ultimately, the raw SNP data is gradually transformed into hybridization schemes that can directly guide breeding, forming a complete and automated closed loop of data input and decision output.
[0068] It should be noted that the above embodiments are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. Those skilled in the art can make other variations or modifications based on the above description. It is impossible to exhaustively list all possible implementations here. All obvious variations or modifications derived from the technical solutions of this invention are still within the scope of protection of this invention.
Claims
1. A high-efficiency and accurate wheat parent matching method based on core germplasm and genome distance, characterized in that, The method comprises the following steps: (1) obtaining whole genome SNP genotype data of a breeding population; (2) constructing a core germplasm with a sample size smaller than the original population and retaining ≥95% of the original genetic diversity based on the whole genome SNP genotype data; (3) analyzing the population genetic structure of the core germplasm and dividing genetic subpopulations; (4) calculating the genomic distance between individuals in the core germplasm; (5) taking the parent from different genetic subpopulations as the primary condition, and taking the genomic distance in the upper quartile to the maximum value interval as the preliminary screening range, further screening the Top N combinations with the maximum and minimum genomic distance in the range, and N is the number of required output combinations; (6) outputting a recommended hybridization combination list.
2. The method for efficient and precise selection of wheat parents based on core germplasm and genome distance according to claim 1, characterized in that, In step (2), the sample size of the core germplasm is 15%-25% of the original population.
3. The method for efficient and precise selection of wheat parents based on core germplasm and genome distance according to claim 1, characterized in that, In step (4), the genomic distance is a normalized Manhattan distance.
4. A wheat parent efficient and accurate matching system, characterized in that, The selection system can realize the selection method of any one of claims 1 to 3, and the selection system comprises the following functional modules: a data acquisition and storage module: the module is responsible for receiving, verifying and storing the original whole genome SNP genotype data transmitted from the external detection platform, and converting it into a standardized data matrix that can be processed by the system; a core germplasm optimization construction module: the module is responsible for receiving the genotype data of the whole population, running the built-in optimization algorithm, and selecting the optimal core germplasm subset from the original population with genetic diversity coverage ≥95% and minimum sample size as the target; a multi-dimensional genetic analysis module: the module performs two analysis tasks in parallel, one of which is a population genetic structure analysis performed by the population structure analysis unit on the core germplasm, dividing genetic subpopulations and assigning subpopulation labels to each sample, and the other is a genomic distance calculation performed by the genomic distance calculation unit between each other in the core germplasm, generating a distance matrix; an intelligent parent selection decision module: the module is responsible for executing the core decision logic of qualitative first and quantitative second, according to the preset rule of first screening cross-subpopulation combinations and then optimizing genomic distance, sorting all possible combinations according to genomic distance, and selecting the optimal parent combination; a hybridization scheme generation and output module: the module is responsible for converting the machine-readable screening results into a scheme that can be directly understood and executed by those skilled in the art, generating a clear and structured recommended hybridization combination list; The above functional modules work together through data flow and control logic, specifically: The data acquisition and storage module is the starting point of the system, and its output is directly input to the core germplasm optimization construction module. After the core germplasm optimization construction module is processed, the generated core germplasm genotype data is output to the multi-dimensional genetic analysis module. After the multi-dimensional genetic analysis module is processed, the population structure division result and the genomic distance matrix are simultaneously output to the intelligent parent selection decision module. After the intelligent parent selection decision module is processed, the final selected optimal parent combination list is output to the hybridization scheme generation and output module. The hybridization scheme generation and output module is the end of the system.