Optimization method for a guide for the DUS testing of new varieties of Crataegus

Abstract

The present application belongs to the field of new plant variety testing, and particularly relates to an optimization method for a Crataegus new variety DUS test guide, comprising the following steps: collecting test materials, determining three types of test traits of quality, false quality and quantity, and obtaining measured codes; combining guide codes and measured codes to calculate the conformity coefficients of standard varieties and test traits, and screening to-be-optimized objects; determining whether to remove the to-be-optimized quality or false quality traits through distribution frequency and uniformity calculation; determining whether to retain the to-be-optimized quantity traits through variation, correlation and principal component analysis, performing normality test on the retained traits, grading the normal or approximately normal traits by using the least significant difference method, and grading the non-normal traits by using the range method; and selecting the varieties with measured values matching the standard values at all levels as new standard varieties. The present application realizes all-around optimization of the DUS test guide, and is very beneficial to the practical application of the guide, plant variety identification, protection and breeding.

Description

Technical Field

[0001] This invention belongs to the field of plant variety testing, specifically relating to an optimized method for DUS testing guidelines for new varieties of hawthorn plants. Background Technology

[0002] New plant varieties are the core competitiveness of the modern seed industry and an important tool for biosecurity and ecological civilization construction. The protection of new plant varieties is a crucial component of the seed industry's revitalization, not only affecting the vital interests of breeders but also playing a key role in promoting innovation and development within the seed industry. Therefore, to protect the legitimate rights of breeders, the granting of rights for new plant varieties must be completed under scientific and standardized authorization criteria. The International Union for the Protection of New Varieties (UPOV) is an international organization established specifically to protect new plant varieties and safeguard the legitimate rights and interests of applicants for plant variety rights. The UPOV Convention it established is the international legal framework for the protection of new plant varieties.

[0003] Hawthorn belongs to the genus Crataegus of the family Rosaceae (Rosaceae). Crataegus Hawthorn (L.) has fruits that can be eaten fresh, but are more commonly used in processed products. It is highly valued for its medicinal properties; hawthorn can promote digestion, maintain cardiovascular health, and enhance nerve function. Currently, its application in landscape design and ecological construction is becoming increasingly widespread, further solidifying its status as an important horticultural plant.

[0004] Currently, DUS testing guidelines have been established for various plants, including those in the *Crataegus* genus, with qualitative traits, pseudo-qualitative traits, and quantitative traits being important components. During testing, the identification of qualitative and pseudo-qualitative traits is relatively easy, while quantitative traits are greatly influenced by environmental and cultivation conditions, exhibiting complex phenotypic variations and thus being more difficult to identify. To correct for the impact of environmental changes and to standardize varietal descriptions, standard varieties are established to represent different expression states of quantitative traits. Therefore, evaluating the accuracy and stability of the standard varieties and tested traits set in the DUS testing guidelines, analyzing and evaluating the tested traits, and developing a grading framework for quantitative traits to facilitate testing practice are of great significance for the practical application and optimization of the testing guidelines, as well as for plant variety identification, conservation, and breeding.

[0005] Currently, no research has proposed a systematic method for optimizing DUS test guidelines. Some studies have suggested establishing quantitative trait classifications to optimize test guidelines, but these only focus on quantitative traits. Therefore, a comprehensive evaluation of the qualitative, pseudo-qualitative, and quantitative traits of DUS test guidelines, followed by optimization recommendations, is more meaningful and crucial for ensuring the accuracy of DUS test guidelines. Summary of the Invention

[0006] This invention aims to provide an optimization method for the DUS testing guidelines of hawthorn plants. It achieves comprehensive optimization of the DUS testing guidelines through three aspects: "evaluation-analysis-optimization". The results are of great benefit to the practical application of the testing guidelines, plant variety identification, protection and breeding.

[0007] To address the above technical problems, the present invention provides the following technical solution: This invention provides an optimized method for the DUS testing guideline for new varieties of hawthorn plants, comprising the following steps: Hawthorn resources covering standard varieties in the DUS testing guidelines for the genus Crataegus were collected as test materials. The test traits of the test materials were measured, and the test codes were obtained. The test traits included qualitative traits, pseudo-qualitative traits, and quantitative traits. Based on the guide code of the standard variety in the DUS test guide and the actual test code, calculate the compliance coefficient of each standard variety and each test trait, and screen out the standard varieties or test traits with compliance coefficients lower than the preset threshold or below the average value as the objects to be optimized. For standard varieties that need to be optimized, varieties with more stable performance and a higher degree of consistency with the trait description will be re-selected as new standard varieties. For qualitative and pseudo-qualitative traits among the test traits to be optimized, their distribution frequency, genetic diversity index, and distribution evenness are calculated to determine whether to remove the trait to retain the fittest trait. For quantitative traits among the test traits to be optimized, variation analysis, correlation analysis, and principal component analysis are performed to determine whether to retain the trait based on the analysis results. Normality tests are performed on the retained quantitative traits. Quantitative traits that conform to or approximately conform to a normal distribution are classified using the least significant difference method, while quantitative traits that do not conform to a normal distribution are classified using the range method, thus establishing a classification framework for quantitative traits. After completing the quantitative trait classification, new codes are obtained based on the new classification intervals. Plant varieties whose measured values ​​of quantitative traits are located at or near the standard values ​​of each level are selected as new standard varieties, replacing standard varieties whose conformity coefficients are lower than the average or at the lowest value.

[0008] As a preferred embodiment of the present invention, the BoxPlot method and the 3σ method are used to verify the validity of the acquired measured data. For data that are determined to be outliers by both methods, the original records or field samples need to be manually checked. If the data input is incorrect, it is corrected directly. If it is an objective fact and the number of outliers does not exceed two, the average of the previous and subsequent measurements is used for correction. If the number of outliers exceeds two, no action is taken, and the relative variance method or the COYU (Combined-Over-Years Uniformity) method is used for consistency verification.

[0009] In a preferred embodiment of the present invention, the formula for calculating the conformity coefficient is as follows: ; Where i represents the conformity coefficient value of a certain standard variety or test trait, a represents the guide code of the standard variety or test trait in the DUS test guide, b represents the measured code of the standard variety or test trait, and X represents the average value of the guide code of the standard variety or test trait.

[0010] In a preferred embodiment of the present invention, the preset threshold is 0.950. When the conformity coefficient is greater than or equal to 0.950, the corresponding standard variety or test trait is determined to be applicable. When the conformity coefficient is less than 0.950, the corresponding standard variety or test trait is determined to be unstable and is used as an object to be optimized.

[0011] In a preferred embodiment of the present invention, the distribution frequency, genetic diversity index, and distribution evenness of qualitative and pseudo-qualitative traits are calculated. The formula for calculating the genetic diversity index is as follows: ; Wherein, H′ represents the genetic diversity index, Pi represents the percentage of varieties in the i-th level of a certain trait, and ln represents the natural logarithm; the distribution evenness is the ratio of the genetic diversity index to the maximum genetic diversity index; when the distribution frequency of a certain trait is greater than 95% or the distribution evenness is less than 0.05, the trait is removed.

[0012] As a preferred embodiment of the present invention, the variation analysis of quantitative traits specifically includes: calculating the intra-variety coefficient of variation and the inter-variety coefficient of variation of the quantitative trait; when the inter-variety coefficient of variation is greater than the intra-variety coefficient of variation, the trait is retained; otherwise, it is discarded.

[0013] As a preferred embodiment of the present invention, correlation analysis and principal component analysis are performed on quantitative traits, specifically including: Calculate the correlation coefficient between each quantitative trait. For trait pairs with a correlation coefficient greater than 0.950, it is determined that there is information redundancy, and one of the traits is deleted. Based on the loading values ​​of each trait in the principal component analysis, the quantitative traits are ranked according to their overall importance. Traits with high loading values ​​are designated as core traits and given more refined grading levels in subsequent classifications, while traits with low loading values ​​are given more coarse grading levels in subsequent classifications.

[0014] As a preferred embodiment of the present invention, the process of classifying quantitative traits that conform to or approximately conform to a normal distribution using the least significant difference method specifically includes: Based on the measurement data of the experimental materials, the mean of the raw data was used as the central standard value, and the least significant difference (LSD) was used as the standard value. 0.05 The value serves as a grade difference, setting standard values ​​for each grade level; The intervals formed by extending 1 / 2 grade difference to both sides of the standard value at each level are used as the grading intervals for each code; if the grading range is less than 3 levels, the trait is removed; if the grading range is greater than 9 levels, the LSD is increased. 0.05 The values ​​are adjusted in multiples to keep the range within 9 levels.

[0015] As a preferred embodiment of the present invention, the process of classifying non-normally distributed quantitative traits using the range method specifically includes: Based on the measurement data of the test materials, the preliminary series is determined by the ratio of the range to the mean of twice the standard deviation. The preliminary series is then rounded to the nearest integer 1, 3, 5, 7, 9 to determine the final series. The ratio of the range to the final series is used as the series difference. The median is used as the center point, and the series boundary is determined by the formula y = G ± (1 / 2 + n) x, where y is the numerical boundary of each series, G is the median, n is 0, 1, 2, 3, 4, and x is the series difference.

[0016] In a preferred embodiment of the present invention, after establishing the grading framework for quantitative traits, one-way ANOVA and Kruskal-Wallis test are used to verify the grading results and confirm the rationality and discriminative power of the grading intervals.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) This invention is the first to propose a comprehensive optimization method covering standard varieties, quality traits, pseudo-quality traits and quantitative traits, breaking through the limitation of existing research that only focuses on quantitative traits. Through an integrated evaluation-analysis-optimization process, the accuracy, stability and practicality of the DUS test guidelines are systematically improved.

[0018] (2) This invention introduces the Coefficient of Conformity (COC) to quantitatively evaluate standard varieties and tested traits, which can accurately identify varieties and traits with unstable expression or poor adaptability, providing a clear basis for subsequent optimization. This method can be extended to the optimization of DUS testing guidelines for other genera and species.

[0019] (3) For quality traits and pseudo-quality traits, quantitative evaluation is carried out using distribution frequency and distribution evenness (genetic diversity index / maximum genetic diversity index) to eliminate traits with poor distinguishing ability or redundant information, thereby improving testing efficiency.

[0020] (4) Combining variation analysis, correlation analysis, and principal component analysis, core traits with strong discriminative power and independent information are retained. Based on the normality test results, the LSD method (normal / approximately normal) and the range method (skewed) are used for grading to ensure the rationality and stability of the grading intervals. This method significantly improves the operability and discrimination accuracy of quantitative traits in the DUS test.

[0021] (5) Based on the new grading intervals, varieties whose measured values ​​match the standard values ​​of each level are selected as new standard varieties, and varieties with low COC or at the average value are replaced first, thereby enhancing the timeliness and regional adaptability of the testing guidelines.

[0022] (6) One-way ANOVA and Kruskal-Wallis test were used to verify the grading results to ensure that the grading intervals have significant discriminative power. p <0.001), providing statistical support for the optimization of the DUS testing guidelines. Attached Figure Description

[0023] Figure 1 Correlation analysis plot. * indicates significant correlation, ** indicates highly significant correlation.

[0024] Figure 2 Frequency distribution histogram.

[0025] Figure 3 : Flowchart of the method of the present invention. Detailed Implementation

[0026] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and preferred embodiments.

[0027] This invention provides an optimized method for the DUS testing guideline for new varieties of hawthorn plants, such as... Figure 3 As shown, it includes the following steps: 1. Conduct field research and collect experimental materials. Hawthorn resources covering standard varieties in the Hawthorn DUS testing guidelines were collected. Healthy hawthorn plants free from pests and diseases were selected. The quality traits and false quality traits of the branches, leaves, flowers and fruits of the plants were observed and recorded by visual inspection, and photographs were taken for subsequent testing and comparison. Vernier calipers and other caliper tools were used to measure quantitative traits. The measurement accuracy was set according to the actual collection situation, and the quantitative trait ratios were calculated.

[0028] 2. Record the raw data The original data of quality traits and pseudo-quality traits were recorded using a horizontal data table. Due to the large number of varieties and traits, only the first 17 varieties and the first 12 traits are listed.

[0029] Table 1: Data on quality traits and pseudo-quality traits The raw data of quantitative traits are recorded using a vertical data table. Due to the large number of varieties and traits, only the leaf trait data of the first two varieties are listed.

[0030] Table 2: Quantitative Trait Data Table 3. Data verification The Box Plot and 3σ methods were used to test the validity of the quantitative trait data. Outliers in the Box Plot and 3σ methods were marked with different colors. Due to the large number of varieties and traits, only the leaf trait data of two varieties are listed.

[0031] Table 3: Data Validation After data validation, for data identified as outliers by both methods, the original records or field samples must be manually checked. If the error is due to data entry, it should be corrected directly. If it is an objective fact, and the number of outliers is very small (e.g., no more than two) and the cause is unknown, the average of the preceding and following measurements can be used for correction. If the number of outliers is large, no action is taken, and the relative variance method or the COYU (Combined-Over-Years Uniformity) method is used for consistency testing.

[0032] 4. Standard varieties and test traits investigation and evaluation Based on the code values ​​of standard varieties in the DUS testing guidelines, calculate the Coefficients of Coincidence (COC) between the standard varieties and the tested traits in the Chinese hawthorn DUS testing guidelines. The calculation formula is as follows:

[0033] i= 1-Σ( ab ) 2 / (X×100); i The COC value represents a specific standard variety (or trait). a The guide code that indicates the standard variety (or trait), b This indicates the measured code for the standard variety (or trait). X This represents the average value of the guide code for that standard variety (or trait).

[0034] According to the formula, the COC value ranges from 0 to 1, and the closer the value is to 1, the more accurate and stable the standard variety (or trait) in the guideline is. A compliance coefficient result greater than or equal to 0.950 indicates that the standard variety in the guideline is reasonable and accurate, while a result less than 0.950 indicates that the variety expression is unstable and a suitable standard variety optimization test guideline needs to be set.

[0035] The performance of standard hawthorn varieties was compared with the guideline descriptions, and the results are shown in Table 4. Analysis of 35 standard hawthorn varieties in the DUS guideline showed an average COC value of 0.998. 27 varieties (77.14%) reached a COC of 1.000, and the measured codes perfectly matched the guideline descriptions, indicating excellent regional adaptability and suitability as standard varieties for DUS testing. Eight hawthorn resources had COCs below 1.000, with three (8.57%) below 0.990: Liaohong, Kaiyuan Ruanzi, and Xiaohuangmianzha, with Xiaohuangmianzha showing the lowest COC (0.980).

[0036] Table 4: COC of Hawthorn Standard Varieties Tested by DUS Table 5: COC of DUS test traits of hawthorn Note: In the table, QN represents quantitative traits, QL represents qualitative traits, and PQ represents pseudo-qualitative traits.

[0037] As shown in Table 5, 43 out of 44 tested traits exhibited a COC value greater than 0.990, accounting for 97.73% of the total. Among them, 37 traits had a COC value of 1.000. Qualitative traits showed the highest COC level, with an average COC value of 1.000, indicating that the actual results were completely consistent with the guideline codes. The average COC value of the 15 pseudo-qualitative traits was 0.998, with 11 of them also reaching 1.000. However, the COC values ​​of fruit shape, peel color, main flesh color, and calyx shape were lower than the average. The average COC value of quantitative traits was 0.998, with lenticel density having the lowest COC value (0.980).

[0038] Based on Tables 4 and 5, the COC values ​​of the results in this embodiment are all greater than 0.950, indicating that the hawthorn DUS testing guidelines studied are highly practical and lay the foundation for trait optimization. However, standard varieties and test traits that are at or below the average value should be prioritized for optimization.

[0039] Regarding the optimization of standard varieties, standard varieties with a COC at or below the average can be re-selected as new standard varieties that exhibit more stable performance and a higher degree of consistency with the trait description. For example, the standard variety with the lowest COC, Xiaohuangmianzha, can be replaced with Jinruyi.

[0040] Regarding the tested traits, this invention is based on COC analysis, and focuses on optimizing 7 traits with COC at or below the average value. At the same time, it combines the testing characteristics of three types of traits: qualitative traits, pseudo-qualitative traits, and quantitative traits, and proposes targeted optimization strategies by using genetic diversity analysis and correlation analysis methods respectively.

[0041] 5. Optimization of quality traits and pseudo-quality traits The trait settings in the DUS testing guidelines are limited by the number and types of known varieties. The standard for evaluating a good set of traits is to differentiate more varieties using the fewest traits. For qualitative and pseudo-qualitative traits, this invention provides genetic diversity analysis to optimize traits.

[0042] Genetic diversity analysis methods include: calculating the code distribution frequency and evenness for qualitative and pseudo-qualitative traits. First, the code distribution frequency is calculated, then the genetic diversity index and the maximum genetic diversity index are calculated, and the evenness is obtained by dividing the two. Based on the trait distribution frequency or evenness, it is determined whether a trait should be removed. For example, if the distribution frequency of a trait is greater than 95% (excluding 95%) or the evenness is less than 0.05 (excluding 0.05), the trait is removed. Furthermore, for qualitative and pseudo-qualitative traits, the lack of intuitive trait descriptions during testing can easily lead to judgment bias. It is recommended to supplement the guidelines with illustrated diagrams to clearly illustrate the expression of each trait, thereby reducing subjective judgment differences among testers.

[0043] Based on the grading standards set by the DUS guidelines for qualitative and pseudo-qualitative traits, their distribution frequency, genetic diversity index (H'), and distribution evenness were calculated to evaluate the traits. The formula for the genetic diversity index (H') is:

[0044] H′ = ΣPi lnPi; pi represents the percentage of varieties in the i-th level of a certain trait out of the total number, and ln represents the natural logarithm.

[0045] Based on frequency distribution, genetic diversity analysis, distribution evenness, and COC, the six traits with a frequency distribution greater than 95% were: plant thorns, zigzag shape of one-year-old branches, pubescence on the upper surface of leaves, flower type, main petal color, and petal folds. The four pseudo-quality traits with a COC below the average were: fruit shape, peel color, main pulp color, and sepal shape. These four traits performed well in the genetic diversity analysis. In the current study, based on considerations of the future breeding potential of hawthorn, these 10 traits were retained.

[0046] Table 6: Frequency distribution of 22 quality traits and pseudo-quality traits of hawthorn 6. Optimization of quantitative trait classification (1) Before analyzing quantitative traits, variation analysis should be performed on the quantitative traits to calculate the maximum, minimum, average, mean, standard deviation (SD), within-varie coefficient of variation, and between-varie coefficient of variation. Only when the between-varie variation is significantly greater than the within-varie variation can the trait effectively distinguish varieties. Traits with insufficient distinguishing ability should be excluded. The results show that the between-varie variation of all 11 quantitative traits is significantly higher than the within-varie coefficient of variation, as shown in Table 7.

[0047] Table 7: Variation of 11 quantitative traits in hawthorn (2) Correlation analysis was performed on the 11 quantitative traits for which variation analysis was completed. For trait pairs with a correlation coefficient greater than 0.950, information redundancy was identified, requiring manual review to determine whether one trait should be deleted. The results showed that among the 11 traits, no trait pair had a correlation coefficient greater than 0.950. Figure 1 ).

[0048] (3) Principal component analysis was performed on the 11 quantitative traits for which correlation analysis was completed. Based on the loading values, the quantitative traits were ranked according to their overall importance. Traits with high loading values ​​were identified as core traits and could be assigned more refined grading levels (e.g., level 7) in subsequent grading. Traits with low loading values ​​could be assigned more coarse grading levels (e.g., level 3-5) in subsequent grading.

[0049] Table 8: Eigenvalues, contribution rates, and eigenvectors of each principal component (4) Normality test of quantitative traits Normality tests were performed on the traits retained after the above analysis: KS (Kolmogorov-Smirnov) normality tests and one-way ANOVA were conducted using Origin software, and kurtosis and skewness values ​​were calculated, and frequency distribution histograms were plotted. Figure 2 ), to screen for traits that conform to or approximately conform to a normal distribution.

[0050] The KS test results revealed three distinct distribution patterns. First, petiole length, blade length, blade width, blade length-to-width ratio, fruit aspect ratio, and floret diameter... p The values ​​are all greater than 0.05, indicating that they conform to a normal distribution. Secondly, the two traits of seed kernel width and seed kernel length-to-width ratio... p The values ​​are below 0.05, but the absolute values ​​of kurtosis and skewness are both less than 1, indicating that it follows an approximately normal distribution. Finally, the fruit transverse diameter, fruit longitudinal diameter, and seed length... p Values ​​below 0.05, and absolute values ​​of kurtosis and skewness both exceeding 1, indicate a skewed distribution (Table 9).

[0051] Table 9: KS Normality Test for Quantitative Traits of Hawthorn (5) Grading of quantitative traits Based on the analysis results of steps (1) to (4), and combined with the measured data of each quantitative trait, the grading interval of each quantitative trait is determined, forming an optimized quantitative trait grading system. The grading of quantitative traits is based on the phenotypic status described in the guidelines, and has been appropriately adjusted according to the actual measurement data. For traits that follow a normal or approximately normal distribution, the LSD method is used; while for traits that deviate from a normal distribution, the range method is used.

[0052] For quantitative traits classified using the LSD method, the classification process is as follows: using the mean of the original data as the central standard value, and using 2 times the LSD... 0.05 The value is used as a grade difference to set standard values ​​for each grade. Extending half a grade difference to both sides of the standard value at each grade level forms the grade interval for each code. The minimum value of each interval is the grade value. Based on this, the grade value and grade interval for each grade are determined. The number and percentage of varieties within each grade interval are counted, and the genetic diversity index is calculated. Based on the statistical results, it is determined whether the coverage of the total interval is between 3 and 9 grades. If it is less than 3 grades, the trait is not suitable for DUS testing and must be excluded; if it is greater than 9 grades, the LSD needs to be increased. 0.05 The grading is adjusted based on multiples of the value, ensuring the grading range is controlled within 9 levels. For traits with a grading range between 3 and 9 levels, 1 to 2 levels can be reserved at each end to accommodate the emergence of new varieties in the future, thus achieving a more reasonable grading. If the minimum value of the minimum grading interval is less than 0, it is set to 0, and then the LSD is determined as described above. 0.05 The multiples are graded from smallest to largest.

[0053] For traits classified using the range method, the number of levels is determined by the formula: Level 1 = Range / (Standard deviation, mean × 2), Level 2 is the smallest value of Level 1 taken from 1, 3, 5, 7, 9, and the range (x) is defined as Range / Level 2. Classification is carried out with the median (G) as the center point, and the classification boundary is obtained by the formula y = G ± (1 / 2 + n) x, where "y" is the numerical boundary of each level, and n = 0, 1, 2, 3, 4.

[0054] Ultimately, eight traits exhibiting a normal or near-normal distribution were classified into five or seven grades: petiole length and seed kernel width were classified into five grades, while the remaining traits were classified into seven grades. In contrast, the three traits deviating from a normal distribution—fruit length, fruit width, and seed kernel length—were all classified into five grades. Based on these grading criteria, the genetic diversity indices derived from the genetic diversity analysis of the 11 quantitative traits ranged from 1.070 to 1.614, with a mean of 1.324.

[0055] Table 10: Grading ranges of quantitative traits conforming to a normal distribution (measurements of leaves and flowers were performed using vernier calipers with an accuracy of 0.1 cm, while measurements of fruits and seeds were performed using vernier calipers with an accuracy of 0.01 cm.)

[0056] Table 11: Range of Quantitative Traits that Do Not Follow a Normal Distribution (6) Verification of quantitative trait grading results The grading results were validated using genetic diversity indicators (see Table 12). All traits showed highly significant differences between different groups. p <0.001). Furthermore, this extreme significance remained constant for traits analyzed using the LSD scoring method ( p <0.001). The F-value is significantly greater than 1 ( p The fact that the inter-group variation was <0.001 confirms that the variation between groups exceeded the random error, thus validating the rationality of the proposed grading scheme. Among the traits evaluated, the flower diameter (F = 195.473) and fruit length-to-width ratio (F = 185.137) had the highest F values, indicating that they have stronger discriminative ability.

[0057] Validation of the grading results based on the range method shows that the results are highly significant. p For traits with a value <0.001, a larger H value indicates a more significant difference between the defined grades, thus confirming the robustness of the grading scheme. Among them, the nucleus length has the highest H value (H = 62.622), which confirms that this trait has a relatively strong distinguishing ability in grading.

[0058] Table 12: Validation of Grading Results (7) Application of quantitative trait grading results After grading is completed, new codes are obtained based on the new grading intervals. These new codes can be used for DUS testing. At the same time, plant varieties whose measured values ​​of quantitative traits are at or near the standard values ​​of each level can be selected as new standard varieties. Compared with the standard varieties in the current guidelines, better standard varieties can be screened out, and standard varieties with lower COC values ​​can be prioritized for optimization.

[0059] The above description is only a preferred embodiment of the present invention. For those skilled in the art, there will be changes in the specific implementation and application scope based on the ideas of the present invention. The content of this specification should not be construed as a limitation of the present invention.

Claims

1. An optimized method for the DUS testing guideline for new varieties of Crataegus pinnatifida, characterized in that, Includes the following steps: Hawthorn resources covering standard varieties in the DUS testing guidelines for the genus Crataegus were collected as test materials. The test traits of the test materials were measured, and the test codes were obtained. The test traits included qualitative traits, pseudo-qualitative traits, and quantitative traits. Based on the guide code of the standard variety in the DUS test guide and the actual test code, calculate the compliance coefficient of each standard variety and each test trait, and screen out the standard varieties or test traits with compliance coefficients lower than the preset threshold or below the average value as the objects to be optimized. For standard varieties that need to be optimized, varieties with more stable performance and a higher degree of consistency with the trait description will be re-selected as new standard varieties. For qualitative and pseudo-qualitative traits among the test traits to be optimized, calculate their distribution frequency and distribution evenness to determine whether to remove the trait to retain the fit trait; for quantitative traits among the test traits to be optimized, perform variation analysis, correlation analysis and principal component analysis, and determine whether to retain the trait based on the analysis results. Normality tests were performed on the retained quantitative traits. The least significant difference method was used to classify the quantitative traits that conform to or approximately conform to a normal distribution, and the range method was used to classify the quantitative traits that do not conform to a normal distribution, thus establishing a classification framework for quantitative traits. After completing the quantitative trait classification, new codes are obtained based on the new classification intervals. Plant varieties whose measured values ​​of quantitative traits are located at or near the standard values ​​of each level are selected as new standard varieties, replacing standard varieties whose conformity coefficients are lower than the average or at the lowest value.

2. The optimization method according to claim 1, characterized in that, The Box Plot method and the 3σ method were used to verify the validity of the acquired measured data. For data that were identified as outliers by both methods, the original records or field samples were manually checked. If the data was an input error, it was corrected directly. If it was an objective fact and the number of outliers did not exceed two, the average of the previous and subsequent measurements was used for correction. If the number of outliers exceeded two, no action was taken, and the relative variance method or the COYU method was used for consistency verification.

3. The optimization method according to claim 1, characterized in that, The formula for calculating the conformity coefficient is: ； Where i represents the conformity coefficient value of a certain standard variety or test trait, a represents the guide code of the standard variety or test trait in the DUS test guide, b represents the measured code of the standard variety or test trait, and X represents the average value of the guide code of the standard variety or test trait.

4. The optimization method according to claim 3, characterized in that, The preset threshold is 0.

950. When the conformity coefficient is greater than or equal to 0.950, the corresponding standard variety or test trait is determined to be applicable. When the conformity coefficient is less than 0.950, the corresponding standard variety or test trait is determined to be unstable and is used as an object to be optimized.

5. The optimization method according to claim 1, characterized in that, The distribution frequency and evenness of qualitative and pseudo-qualitative traits are calculated. The formula for calculating the genetic diversity index is as follows: ； Wherein, H′ represents the genetic diversity index, Pi represents the percentage of varieties in the i-th level of a certain trait, and ln represents the natural logarithm; the distribution evenness is the ratio of the genetic diversity index to the maximum genetic diversity index; when the distribution frequency of a certain trait is greater than 95% or the distribution evenness is less than 0.05, the trait is removed.

6. The optimization method according to claim 1, characterized in that, Variation analysis of quantitative traits includes calculating the intra-varie coefficient of variation and the inter-varie coefficient of variation of the quantitative trait. If the inter-varie coefficient of variation is greater than the intra-varie coefficient of variation, the trait is retained; otherwise, it is discarded.

7. The optimization method according to claim 1, characterized in that, Correlation analysis and principal component analysis were performed on quantitative traits, specifically including: Calculate the correlation coefficient between each quantitative trait. For trait pairs with a correlation coefficient greater than 0.950, it is determined that there is information redundancy, and one of the traits is deleted. Based on the loading values ​​of each trait in the principal component analysis, the quantitative traits are ranked according to their overall importance. Traits with high loading values ​​are designated as core traits and given more refined grading levels in subsequent classifications, while traits with low loading values ​​are given more coarse grading levels in subsequent classifications.

8. The optimization method according to claim 1, characterized in that, The process of classifying quantitative traits that conform to or approximately conform to a normal distribution using the least significant difference method specifically includes: Based on the measurement data of the experimental materials, the mean of the raw data was used as the central standard value, and the least significant difference (LSD) was used as the standard value. 0.05 The value serves as a grade difference, setting standard values ​​for each grade level; The intervals formed by extending 1 / 2 grade difference to both sides of the standard value at each level are used as the grading intervals for each code; if the grading range is less than 3 levels, the trait is removed; if the grading range is greater than 9 levels, the LSD is increased. 0.05 The values ​​are adjusted in multiples to keep the range within 9 levels.

9. The optimization method according to claim 1, characterized in that, The process of classifying non-normally distributed quantitative traits using the range method specifically includes: Based on the measurement data of the test materials, the preliminary series is determined by the ratio of the range to the mean of twice the standard deviation. The preliminary series is then rounded to the nearest integer 1, 3, 5, 7, 9 to determine the final series. The ratio of the range to the final series is used as the series difference. The median is used as the center point, and the series boundary is determined by the formula y = G ± (1 / 2 +n) x, where y is the numerical boundary of each series, G is the median, n is 0, 1, 2, 3, 4, and x is the series difference.

10. The optimization method according to any one of claims 1 to 9, characterized in that, After establishing the grading framework for quantitative traits, one-way ANOVA and Kruskal-Wallis test were used to verify the grading results and confirm the rationality and discriminative power of the grading intervals.

Images