A method of germplasm screening

Abstract

The application provides a germplasm screening method, comprising the following steps: obtaining a preliminary screening sample set containing multi-phenotype data; performing normalization processing on the phenotype values to eliminate heterogeneity between traits; calculating a comprehensive score of each sample based on the normalized values; and finally screening out a preferred sample set according to the score. Through the automatic normalization and comprehensive scoring process, the problem that multi-trait data is difficult to integrate due to different measurement scales and genetic backgrounds is effectively solved, the breeder is liberated from tedious and subjective manual judgment, a standard and efficient large-scale germplasm screening scheme is formed, and the accuracy and efficiency of excellent germplasm resource mining are significantly improved.

Description

Technical Field

[0001] This application relates to the field of germplasm screening technology, specifically to a germplasm screening method. Background Technology

[0002] In crop genetic improvement and breeding practices, the discovery and evaluation of superior germplasm resources is a fundamental step. As breeding goals have shifted from optimizing a single trait to emphasizing multiple traits such as yield, resistance, quality, and adaptability, traditional screening methods relying on single-trait performance are no longer sufficient to meet practical needs. Especially in the context of modern agriculture, breeders need to comprehensively evaluate and weigh multiple complex traits that may involve trade-offs in order to select materials that possess advantages in multiple aspects.

[0003] In existing technologies, the screening of multi-trait germplasm often relies on breeders manually setting thresholds and independently evaluating each trait, or performing simple subjective weighted comparisons. Different traits have significantly different genetic backgrounds, environmental sensitivities, and measurement scales, making it difficult for traditional methods to account for the interactions and trade-offs between traits. Summary of the Invention

[0004] In view of the above-mentioned defects or deficiencies in the prior art, it is desirable to provide a germplasm screening method, including the following steps:

[0005] Obtain a preliminary screening sample set, which includes the sample number of all preliminary screening samples and their corresponding datasets. Each dataset includes multiple phenotypes and phenotype values ​​corresponding to each phenotype.

[0006] Based on all phenotypic values ​​corresponding to the same phenotypic in the preliminary screening sample set, the phenotypic values ​​corresponding to the phenotypic in each preliminary screening sample are normalized to obtain the normalized phenotypic values ​​of each phenotypic in each preliminary screening sample.

[0007] A comprehensive score set is calculated based on the normalized phenotypic values. The comprehensive score set includes each of the preliminary screening samples and its corresponding comprehensive score value.

[0008] The comprehensive score set is filtered to obtain the preferred sample set, which includes the sample number of multiple preferred samples and their corresponding datasets.

[0009] According to the technical solution provided in this application, obtaining the preliminary screening sample set includes the following steps:

[0010] Obtain an initial sample set, which includes the sample number of all initial samples and their corresponding dataset. The dataset includes multiple phenotypes and phenotype values ​​corresponding to each phenotype.

[0011] Obtain the filtering conditions for each phenotype, the filtering conditions including basic conditions;

[0012] Determine whether the phenotype satisfies the basic conditions corresponding to the phenotype;

[0013] If the conditions are met, then the corresponding initial sample is set as the preliminary screening sample;

[0014] The preliminary screening sample set is obtained based on the preliminary screening samples and their corresponding datasets.

[0015] According to the technical solution provided in this application, the process of filtering the comprehensive score set to obtain a preferred sample set, wherein the preferred sample set includes the sample number of multiple preferred samples and their corresponding dataset, includes the following steps:

[0016] Set a percentage of high-quality products;

[0017] The preliminary screening samples in the comprehensive score set are sorted according to the comprehensive score value.

[0018] Based on the stated percentage of high quality, the top-ranked preliminary screening samples are selected as candidate samples, and a candidate sample set is constructed, which includes each candidate sample and its corresponding dataset.

[0019] The candidate sample set is filtered to obtain the preferred sample set.

[0020] According to the technical solution provided in this application, the screening conditions further include quality conditions, and the screening of the candidate sample set to obtain the preferred sample set includes the following steps:

[0021] Determine whether each phenotype of the candidate sample satisfies the corresponding quality condition;

[0022] If at least one obtained phenotype satisfies the quality condition, then the candidate sample corresponding to it is set as the preferred sample;

[0023] Construct the preferred sample set, which includes the preferred samples and their corresponding datasets.

[0024] According to the technical solution provided in this application, the step of filtering the comprehensive score set to obtain a preferred sample set, wherein the preferred sample set includes the sample number of multiple preferred samples and their corresponding datasets, further includes the following steps:

[0025] Obtain the phenotypic level of each phenotype in the preferred sample, wherein the phenotypic level includes qualified and excellent;

[0026] A dynamic interaction graph is constructed based on the preferred sample set. The dynamic interaction graph intuitively displays the sample number, phenotype, phenotype value corresponding to each phenotype, and phenotype level corresponding to each phenotype for each preferred sample.

[0027] According to the technical solution provided in this application, the phenotypic level of each phenotype in the preferred sample is obtained in the following manner:

[0028] The phenotypic level of the phenotypes in the preferred samples that meet the high-quality criteria is set to high-quality, and the phenotypic level of the phenotypes in the preferred samples that do not meet the high-quality criteria is set to qualified.

[0029] According to the technical solution provided in this application, the comprehensive score set is calculated based on the normalized phenotypic values. The comprehensive score set includes each of the preliminary screened samples and its corresponding comprehensive score value, which is achieved by the following formula:

[0030] Formula (1)

[0031] in, Score This represents the overall score. i Represents the first in the dataset i Each phenotype n This represents the number of phenotypes in the dataset. Representing the i The normalized phenotypic value corresponding to each phenotypic. W i Representing the i Any value within the range of phenotypic weights corresponding to a phenotypic. T i Representing the i Phenotypic trend coefficients corresponding to each phenotype.

[0032] According to the technical solution provided in this application, after calculating the comprehensive score set based on the normalized phenotypic value, all the comprehensive score values ​​need to be corrected before filtering the comprehensive score set so that the comprehensive score values ​​are positive.

[0033] According to the technical solution provided in this application, based on all phenotypic values ​​corresponding to the same phenotypic in the preliminary screening sample set, the phenotypic values ​​corresponding to the phenotypic in each preliminary screening sample are normalized to obtain the normalized phenotypic values ​​of each phenotypic in each preliminary screening sample, which are obtained by the following formula:

[0034] Formula (2)

[0035] in, phe i Represents the first in a certain preliminary screening samplei Phenotypic values ​​of a phenotype phe min This represents the smallest phenotypic value among all initially screened samples for this phenotypic type. phe max This represents the maximum phenotypic value of the phenotype among all initially screened samples.

[0036] Compared with existing technologies, the advantages of this application are as follows: By introducing an automated normalization process, this invention effectively eliminates the data heterogeneity problem caused by differences in genetic background, environmental sensitivity, and measurement scales among different phenotypic traits, providing a unified and comparable benchmark for the integration of multi-trait data. Based on this, by comprehensively scoring the normalized phenotypic values, it achieves the quantification and comprehensive evaluation of complex trade-offs among multiple traits, thereby freeing breeders from tedious and subjective manual judgment. This method forms a standardized and efficient screening process, enabling rapid and objective identification of materials with comprehensive trait advantages that excel in multiple traits from large-scale germplasm resources, greatly improving the accuracy and efficiency of screening. Attached Figure Description

[0037] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0038] Figure 1 A flowchart illustrating a germplasm screening method provided in this application embodiment;

[0039] Figure 2 Dynamic interactive diagrams provided for embodiments of this application; Detailed Implementation

[0040] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0041] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0042] As mentioned in the background section, this application proposes a germplasm screening method, such as... Figure 1 As shown, it includes the following steps:

[0043] S100. Obtain a preliminary screening sample set, which includes the sample number of all preliminary screening samples and their corresponding datasets. Each dataset includes multiple phenotypes and phenotype values ​​corresponding to each phenotype.

[0044] The process of obtaining the initial screening sample set includes the following steps:

[0045] S110. Obtain an initial sample set, the initial sample set including the sample number of all initial samples and the corresponding dataset, the dataset including multiple phenotypes and phenotype values ​​corresponding to each phenotype;

[0046] The initial sample set refers to the original germplasm resource data that has not undergone any screening, and may optionally be derived from field trials, gene banks, or public databases. In this embodiment, 1047 initial samples were obtained by hybridizing three widely planted superior wheat varieties, Zhoumai 18 (high-yield), Zhengmai 366 (improved variety), and Handan 6172, as well as samples of major domestic backbone parents.

[0047] Each initial sample has a unique sample number and is associated with a dataset containing multiple phenotypes and their phenotypic values. For example, phenotypes in field planting records include heading date (HD, days), flowering date (FD, days), thousand-grain weight (TGW, grams), and resistance to leaf rust (Lr) and stripe rust (Yr). Disease severity is scored from 0 to 4, representing different responses from immunity to severe susceptibility. Phenotypic values ​​are specific measurements (e.g., HD is 180 days, TGW is 50 grams).

[0048] S120. Obtain the filtering conditions for each phenotype, wherein the filtering conditions include basic conditions;

[0049] The baseline conditions represent the minimum acceptance criteria for each phenotype. For example, baseline conditions can be defined as: FD: 181; HD: 181; Lr: <4; TGW: 42; Yr: <4. The super threshold is defined as: FD: >=189; HD: >=185; Lr: <=2; TGW: >48; Yr: <=2. Baseline conditions are usually set by breeding experts based on experience or industry standards, reflecting the basic requirements of germplasm resources.

[0050] S130. Determine whether the phenotype satisfies the basic conditions corresponding to the phenotype;

[0051] S140. If satisfied, the initial sample corresponding to it is set as the preliminary screening sample;

[0052] Specifically, each initial sample in the initial sample set is traversed, and each phenotypic value is checked to see if it meets the corresponding basic conditions. For example, for initial sample A, if its FD=185, TGW=45, and Lr=3, then all phenotypic values ​​meet the basic conditions; if a sample has TGW=40, then it does not meet the conditions and will be excluded.

[0053] S150. Obtain the preliminary screening sample set based on the preliminary screening samples and their corresponding datasets.

[0054] Specifically, an initial screening sample set is constructed based on the initial screening samples. This set contains only the initial samples that have been screened through the basic criteria and their complete datasets.

[0055] This application, based on the principle of threshold filtering, rapidly eliminates substandard samples by setting scientific baseline conditions, reducing subsequent computational load. It simulates a common initial screening process in breeding, ensuring that only samples with potentially superior traits proceed to in-depth analysis. Automated initial screening avoids the tediousness and errors of manual inspection, making it particularly suitable for large-scale germplasm banks. Flexible setting of baseline conditions allows for customized screening criteria based on different breeding objectives, enhancing the method's applicability and improving the efficiency and relevance of the screening process.

[0056] S200. Based on all phenotypic values ​​corresponding to the same phenotypic in the preliminary screening sample set, normalize the phenotypic values ​​corresponding to the phenotypic in each preliminary screening sample to obtain the normalized phenotypic values ​​for each phenotypic in each preliminary screening sample; obtained through the following formula:

[0057] Formula (2)

[0058] in, phe i Represents the first in a certain preliminary screening sample i Phenotypic values ​​of a phenotype phe min This represents the smallest phenotypic value among all initially screened samples for this phenotypic type. phe max This represents the maximum phenotypic value for that phenotype among all initially screened samples. Representing the i The normalized phenotypic value corresponding to each phenotypic.

[0059] Taking a preliminary screening set containing three samples (sample numbers A, B, and C), and focusing on two phenotypes: thousand-grain weight and flowering time, as an example, some data from the preliminary screening set are shown in Table 1:

[0060] Table 1

[0061]

[0062] For the phenotypic thousand-grain weight, among all the preliminary screening samples, the minimum phenotypic value was 45 grams and the maximum phenotypic value was 55 grams. The normalized phenotypic value of the thousand-grain weight of the preliminary screening sample with sample number A was: (45-45) / (55-45)=0; the normalized phenotypic value at the flowering stage was: (185-181) / (189-181)=0.5

[0063] Normalization linearly maps the original data to the [0,1] interval, eliminating dimensions and unifying the "goodness" of all traits as relative scores between 0 and 1, where 1 represents the best performance of the trait in the current population and 0 represents the worst. This lays the mathematical foundation for applying different weights to different traits. After normalization, breeders do not need to worry that a certain trait (such as thousand-grain weight, usually in the tens of grams) will naturally dominate the overall score due to its large absolute value, while another trait (such as disease score, between 0 and 4) will be insignificant due to its small absolute value. Normalization ensures that the contribution of each trait is only related to its weight when calculating the overall score, thus greatly improving the fairness, accuracy, and scientific rigor of multi-trait screening.

[0064] S300. A comprehensive score set is calculated based on the normalized phenotypic values. The comprehensive score set includes each of the initially screened samples and its corresponding comprehensive score value; this is achieved through the following formula:

[0065] Formula (1)

[0066] in, Score This represents the overall score. i Represents the first in the dataset i Each phenotype n This represents the number of phenotypes in the dataset. Representing the i The normalized phenotypic value corresponding to each phenotypic. W i Representing the i Any value within the range of phenotypic weights corresponding to a phenotypic. T i Representing the i Phenotypic trend coefficients corresponding to each phenotype.

[0067] Specifically, the phenotypic weight range represents the variable range of importance for each phenotypic trait. This weight range is set by breeding experts and reflects the relative importance of the trait. The phenotypic trend coefficient indicates the optimization direction of the phenotypic trait: 1 indicates a higher value is better (e.g., yield), and -1 indicates a lower value is better (e.g., disease score). The phenotypic trend coefficient ensures the correctness of the calculation direction. Each initial screening sample contains the same phenotypic type, and the phenotypic weight range and phenotypic trend coefficient are the same for the same phenotypic traits.

[0068] The phenotypic weight range and the definition of the phenotypic trend coefficient are shown in Table 2:

[0069] Table 2

[0070]

[0071] Where H is [0.8, 1.0], D is [0.4, 0.6], and Z is [0.1, 0.3]. Taking sample A in Table 1 as an example, the phenotypic weight of 1,000 grains ranges from [0.8, 1.0]. Any value can be randomly selected from this range. W 1 = 0.9, because the thousand-grain weight is "the higher the better," and its phenotypic trend coefficient is... T =1; The importance of the flowering period phenotype is relatively low, and the phenotype weight ranges from [0.4, 0.6], from which any value can be selected. W 2 = 0.5, because the expected value is "early maturity" (fewer flowering days), so the smaller the phenotypic trend coefficient, the better. T 2 = -1; therefore, according to formula (1), the comprehensive score of sample A = 0 * 0.9 * 1 + 0.5 * 0.5 * (-1) = -0.25.

[0072] Formula (1) is based on a weighted summation model, a classic method for multi-attribute decision-making. Phenotypic weights reflect trait priority, phenotypic trend coefficients control direction, and normalized values ​​ensure fair comparison. This application provides an objective and customizable scoring mechanism to adapt to different breeding objectives. The comprehensive score integrates multi-trait information, reducing the arbitrariness of manual evaluation.

[0073] In a preferred embodiment, after step S300 and before step S400, all the comprehensive score values ​​are corrected to make the comprehensive score values ​​positive.

[0074] Specifically, since the comprehensive score value may be negative because phenotypes with a trend coefficient of -1 (such as diseases) contribute negative values, and negative values ​​are inconvenient for comparison and ranking, correction is required. This correction can be performed using the following formula: Corrected comprehensive score value = 500 - (-comprehensive score value). For example, as mentioned above, the comprehensive score value of sample A is -0.25, and the corrected comprehensive score value = 500 - (-0.25) = 500.25, converting the negative value to a positive value. This application maps the data to the positive domain through correction, making the corrected score easier to process and avoiding the confusion caused by negative values. Simultaneously, it ensures the smooth progress of the screening step, as ranking and percentage calculations typically assume positive values. In this embodiment, the corrected data is shown in Table 3 below (partial data):

[0075] Table 3

[0076]

[0077] S400. Filter the comprehensive score set to obtain a preferred sample set, which includes the sample index of multiple preferred samples and their corresponding datasets. This includes the following steps:

[0078] S410. Set the percentage of superior quality; optionally, the percentage of superior quality is set to 20%. This parameter reflects the breeder's desired proportion of superior germplasm and is usually set based on historical data or expert experience.

[0079] S420. Sort the preliminary screened samples in the comprehensive score set according to the comprehensive score value;

[0080] S430. Based on the stated percentage of high-quality samples, the top-ranked preliminary screening samples are selected as candidate samples, and a candidate sample set is constructed. The candidate sample set includes each candidate sample and its corresponding dataset. This introduces a quantitative screening mechanism based on the "percentage of high-quality samples," which is a proportional screening method. This ensures that the selected samples are at the top level within the overall sample population, avoiding selection biases that might arise from differences in the absolute score threshold due to variations in the sample group. The effect is to provide a fast and objective initial screening method that can efficiently narrow down the focus from a large number of samples, concentrating resources on in-depth analysis of the most promising samples.

[0081] S440. Filter the candidate sample set to obtain the preferred sample set.

[0082] The screening criteria also include quality criteria, and the process of screening the candidate sample set to obtain the preferred sample set includes the following steps:

[0083] S441. Determine whether each phenotype of the candidate sample meets the corresponding quality conditions; for example, the quality conditions are: flowering period (FD): >= 189 days, heading period (HD): >= 185 days, leaf rust resistance (Lr): <= 2 points, thousand-grain weight (TGW): > 48 grams, stripe rust resistance (Yr): <= 2 points;

[0084] S442. If at least one obtained phenotype satisfies the quality condition, then the candidate sample corresponding to it is set as the preferred sample;

[0085] Taking candidate sample number 1066 as an example, FD: 186.62 < 189 (not satisfied), HD: 181.11 < 185 (not satisfied), Lr: 1.37 <= 2 (satisfied), TGW: 55.41 > 48 (satisfied), Yr: 2 <= 2 (satisfied). Candidate sample 1066 meets the high-quality standard in the three traits of Lr, TGW, and Yr, and is marked as a preferred sample.

[0086] S443. Construct the preferred sample set, which includes the preferred samples and their corresponding datasets. The preferred samples in the candidate sample set are not only those with high comprehensive scores, but also germplasm resources that have reached an "excellent" level in at least one key breeding trait. This ensures that the finally selected germplasm resources are not only comprehensively developed, but also highly likely to become valuable breeding parents in specific target traits, greatly improving the breeding application value of the screening results.

[0087] In a preferred embodiment, the following steps are included after step S400:

[0088] S500. Obtain the phenotypic level of each phenotype in the preferred sample, wherein the phenotypic level includes qualified and excellent;

[0089] The phenotypic level of each phenotype in the preferred sample is obtained in the following manner:

[0090] The phenotypic level of the phenotypes in the preferred samples that meet the high-quality criteria is set to "high-quality," and the phenotypic level of the phenotypes in the preferred samples that do not meet the high-quality criteria is set to "qualified." Taking the high-quality sample with sample number 1066 as an example again: FD's level is "qualified," HD's level is "qualified," Lr's level is "high-quality," TGW's level is "high-quality," and Yr's level is "high-quality."

[0091] S600. Construct a dynamic interaction graph based on the preferred sample set. The dynamic interaction graph visually displays the sample number, phenotype, phenotype value corresponding to each phenotype, and phenotype level corresponding to each phenotype for each preferred sample.

[0092] like Figure 2 As shown, each circle represents the same phenotype, and each dot represents a preferred sample. When the mouse moves over any preferred sample, the sample number, phenotype, phenotype value, and level of the preferred sample will be displayed. In addition, the levels of "qualified" and "excellent" can be distinguished by color.

[0093] This application effectively solves three major technical challenges in multi-trait germplasm screening: low efficiency, difficulty in integrating trait heterogeneity, and poor interpretability of results. Its core principle lies in normalizing trait data from different dimensions and genetic backgrounds into comparable indicators, then assigning weights according to breeding objectives for comprehensive evaluation, and finally presenting the results intuitively through a dynamic interactive graph. This achieves a leap from traditional experience-based manual screening to automated, standardized, and visualized screening, significantly improving the efficiency, accuracy, and decision support capabilities of large-scale germplasm resource analysis.

[0094] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A germplasm screening method, characterized in that, Includes the following steps: Obtain a preliminary screening sample set, which includes the sample number of all preliminary screening samples and their corresponding datasets. Each dataset includes multiple phenotypes and phenotype values ​​corresponding to each phenotype. Based on all phenotypic values ​​corresponding to the same phenotypic in the preliminary screening sample set, the phenotypic values ​​corresponding to each phenotypic in each preliminary screening sample are normalized to obtain the normalized phenotypic values ​​of each phenotypic in each preliminary screening sample. A comprehensive score set is calculated based on the normalized phenotypic values. The comprehensive score set includes each of the preliminary screening samples and its corresponding comprehensive score value. The comprehensive score set is filtered to obtain an optimal sample set, which includes the sample number of multiple optimal samples and their corresponding datasets; The process of obtaining the initial screening sample set includes the following steps: Obtain an initial sample set, which includes the sample number of all initial samples and their corresponding dataset. The dataset includes multiple phenotypes and phenotype values ​​corresponding to each phenotype. Obtain the filtering conditions for each phenotype, the filtering conditions including basic conditions; Determine whether each of the phenotypes simultaneously satisfies the basic conditions corresponding to that phenotype; If the conditions are met, then the corresponding initial sample is set as the preliminary screening sample; The preliminary screening sample set is obtained based on the preliminary screening samples and their corresponding datasets; The selection of the comprehensive score set yields a preferred sample set, which includes the sample numbers of multiple preferred samples and their corresponding datasets, and includes the following steps: Set a percentage of high-quality products; The preliminary screening samples in the comprehensive score set are sorted according to the comprehensive score value. Based on the stated percentage of high quality, the top-ranked preliminary screening samples are selected as candidate samples, and a candidate sample set is constructed, which includes each candidate sample and its corresponding dataset. The candidate sample set is filtered to obtain the preferred sample set; The screening criteria also include quality criteria, and the screening of the candidate sample set to obtain the preferred sample set includes the following steps: Determine whether each phenotype of the candidate sample satisfies the corresponding quality condition; If at least one obtained phenotype satisfies the quality condition, then the candidate sample corresponding to it is set as the preferred sample; Constructing the preferred sample set, which includes the preferred samples and their corresponding datasets; the preferred samples are those that simultaneously meet the criteria of high comprehensive score ranking and at least one phenotype meeting the high-quality condition; filtering the comprehensive score set to obtain the preferred sample set, which includes the sample numbers of multiple preferred samples and their corresponding datasets, and then further including the following steps: Obtain the phenotypic level of each phenotype in the preferred sample, wherein the phenotypic level includes qualified and excellent; A dynamic interaction graph is constructed based on the preferred sample set. The dynamic interaction graph intuitively displays the sample number, phenotype, phenotype value corresponding to each phenotype, and phenotype level corresponding to each phenotype for each preferred sample.

2. The germplasm screening method according to claim 1, characterized in that, The phenotypic level of each phenotype in the preferred sample is obtained in the following manner: The phenotypic level of the phenotypes in the preferred samples that meet the high-quality criteria is set to high-quality, and the phenotypic level of the phenotypes in the preferred samples that do not meet the high-quality criteria is set to qualified.

3. The germplasm screening method according to claim 1, characterized in that, The comprehensive score set is calculated based on the normalized phenotypic values. The comprehensive score set includes each of the initially screened samples and its corresponding comprehensive score value, which is achieved through the following formula: Official (1) in, Score This represents the overall score. i Represents the first in the dataset i Each phenotype n This represents the number of phenotypes in the dataset. Representing the i The normalized phenotypic value corresponding to each phenotypic. W i Representing the i Any value within the range of phenotypic weights corresponding to a phenotypic. T i Representing the i Phenotypic trend coefficients corresponding to each phenotype.

4. The germplasm screening method according to claim 1, characterized in that, After calculating the comprehensive score set based on the normalized phenotypic values, all the comprehensive score values ​​need to be corrected before filtering the comprehensive score set to make the comprehensive score values ​​positive.

5. The germplasm screening method according to claim 3, characterized in that, Based on all phenotypic values ​​corresponding to the same phenotypic in the preliminary screening sample set, the phenotypic values ​​corresponding to each phenotypic in each preliminary screening sample are normalized to obtain the normalized phenotypic values ​​of each phenotypic in each preliminary screening sample, which are obtained by the following formula: Official (2) in, phe i Represents the first in a certain preliminary screening sample i Phenotypic values ​​of a phenotype phe min This represents the smallest phenotypic value among all the described phenotypic values ​​in all the initially screened samples. phe max This represents the maximum phenotypic value among all the phenotypes described in the initial screening samples.

Images