CN111460777B - A DUS testing method for plant varieties - Google Patents

Abstract

A DUS testing method for plant varieties, including conducting plant cultivation tests, collecting and processing test data in a unified format, and performing DUS analysis. The method optimizes the DUS test design and increases the objectivity of the DUS test results. At the same time, data and statistical hypothesis validity testing and DUS data analysis can be efficiently implemented with the help of the EXCEL program.

Description

A DUS testing method for plant varieties

Technical Field

The present invention belongs to the technical field of DUS testing of plant varieties, and mainly aims at processing and analyzing data obtained in DUS testing, realizing DUS determination, and optimizing DUS testing guidelines and experimental design methods.

Background Art

Plant variety specificity (Distinctness), consistency (Uniformity) and stability (Stability) test (DUS test for short) refers to the process of evaluating the specificity, consistency and stability of plant varieties through planting tests or indoor analysis using corresponding testing technologies and standards. DUS test is the basic technical basis for variety management in various countries and a necessary condition for the protection, variety approval or registration of new plant varieties. It is of great significance for plant selection and breeding, promoting seed engineering construction, and promoting the development of agricultural and forestry production.

Specificity refers to one or more traits of a plant variety that are clearly distinguishable from known varieties. Uniformity refers to the consistent expression of related characteristics or traits between individuals in a plant variety, in addition to expected natural variation. Stability refers to the fact that the main traits of a plant variety remain unchanged after repeated reproduction or at the end of a specific reproduction cycle.

Internationally, the history of DUS testing is synchronized with the history of the plant variety protection system. In February 1957, the French government invited 12 Western European countries to participate in the diplomatic conference held in Paris in May of the same year to discuss the establishment of a special plant variety protection system. At the second diplomatic conference held in Paris at the end of 1961, the "International Convention for the Protection of New Varieties of Plants" was adopted, and the "International Union for the Protection of New Varieties of Plants (UPOV)" was established on this basis. The convention was revised three times in 1972, 1978 and 1991, forming the most complete DUS terminology and definition. UPOV summarizes the DUS testing experience of its members, organizes the drafting and gradual adoption of 15 TGP technical documents and 323 DUS testing guidelines, and comprehensively standardizes the basic concepts and principles of DUS testing, the confirmation of known varieties and the construction and maintenance of variety libraries, testing experience and cooperation, the development of DUS testing guidelines, experimental design and statistical analysis, DUS review procedures, new plant type testing guidance, and molecular technology application principles.

On March 20, 1997, China promulgated the Regulations on the Protection of New Plant Varieties, which listed DUS as a necessary and substantial condition for variety authorization. The entire regulations were formulated with reference to the 1978 text of UPOV, and the definition of DUS was formulated with reference to the 1991 text of UPOV. On April 23, 1999, China joined the 1978 text of the UPOV Convention and became the 39th member of UPOV. In the same year, it began to accept applications for new plant variety rights. On September 29, 2000, the former Ministry of Agriculture established a DUS testing center and 14 sub-centers across the country to specifically undertake the DUS testing of new plant varieties. On November 4, 2015, the newly revised Seed Law included the protection of new plant varieties in the scope of variety management for the first time, and stipulated that DUS testing was a prerequisite for variety approval and registration. With the continuous increase in the workload of DUS testing, the former Ministry of Agriculture later established 13 sub-centers and 3 testing stations. Up to now, the Ministry of Agriculture and Rural Affairs has released a total of 159 plant protection catalogs, included five crops such as corn in the variety approval catalog, and included 29 crops such as potatoes in the variety registration catalog.

Although UPOV has released a complete set of DUS test guidance documents, it does not provide specific operation methods. The UK's DUST statistical analysis software and France's GAIA variety management software are currently the only two softwares that are shared free of charge within UPOV. However, the former only involves mathematical and statistical analysis functions and the operation is cumbersome, while the latter only involves variety description analysis functions and the weight setting is highly subjective. Both have limitations in use.

In recent years, representatives of various countries have shared their own data analysis methods at the annual meeting of the UPOV Technical Working Group, but they mainly focus on statistical analysis, with scattered functions and many operating steps. In particular, there is a lack of quality control methods for raw data, resulting in unstable analysis results and low analysis efficiency.

The inventors have been continuously conducting in-depth research and have invented a comprehensive DUS analysis method. With only a set of fixed trait parameters, data from different locations and years can be summarized and analyzed to obtain scientific and accurate test results, while the analysis speed is greatly accelerated.

Summary of the invention

In order to efficiently analyze the DUS test data of plant varieties, improve the experimental level and evaluation quality of DUS tests, and provide a comprehensive and accurate basis for DUS determination, the inventors have developed a set of plant variety DUS test experiments and data evaluation and analysis methods. In addition, the data analysis method developed by the inventors can be conveniently implemented with the help of EXCEL program.

In view of this, a plant variety DUS test method is provided, comprising conducting a plant planting test, collecting and processing test data in a unified format, and performing DUS analysis, wherein the variety types in the plant planting test include a variety to be tested, a standard variety, and an optional similar variety; characterized in that collecting and processing the test data in a unified format comprises: making a unified table, setting a unified data format (such as a data type), a numerical range, and an optional numerical unit, and testing the validity of the test data and statistical hypotheses;

The verification of the validity of the test data includes: performing a data format, numerical range and/or optional numerical unit matching test, and the test is automatically performed on the data during or after input with the help of a program. If the input data does not match the set data format, numerical range and/or optional numerical unit, it indicates that an abnormal value has occurred. If an abnormal value has occurred, the abnormal value is automatically marked, and the original record or field sample is manually checked; if it is an input error, it is directly corrected; if it is an objective fact, the abnormal data is retained and the subsequent steps are continued.

The plant planting test may only be conducted in one phase. Preferably, the plant planting test is conducted in two or more phases. A complete reproductive cycle for annual or biennial plants is the time from sowing to harvesting, and for perennial plants it is the time from germination to harvesting in the normal flowering and fruiting year. Plant growth is greatly affected by temperature, rainfall and light, and there will be large differences between years. One year often cannot accurately determine the variety description and differences. Therefore, the DUS test generally requires two growth cycles. For crops or varieties with poor consistency and small differences between varieties (such as forage), 3 growth cycles are often required. For those crops or varieties that are asexually propagated, planted in a controlled environment (greenhouse), and have obvious variety differences (such as Phalaenopsis), 1 growth cycle can also end the DUS test.

Stability testing generally involves planting test analysis of seeds of the same variety from different generations. If the expression of traits of the next generation of seeds is consistent with that of the previous generation, and they are both consistent, it means that the variety is stable. If a variety is consistent in a certain test, it means that it is stable.

The consistency and stability of hybrid varieties can also be judged by testing the consistency or stability of the parents.

Known varieties refer to plant varieties that have been accepted for application or have passed variety evaluation, variety registration, new variety protection, or have been sold and promoted. Varieties to be tested refer to varieties that have applied for variety protection, evaluation or registration, or varieties sampled from the market for evaluation. Similar varieties: refers to varieties that are screened from the variety library for specific testing and are similar to the varieties to be tested in phenotype or molecular characteristics and need to be further verified in field planting trials. Standard varieties: refers to known varieties used in planting trials to evaluate environmental impacts and indicate the expression status of traits.

According to the expression type, traits can be divided into quality traits (QL), pseudo-quality traits (PQ) and quantitative traits (QN). According to the observation type, they can be divided into visual observation (V) and measurement (M). According to the record type, they can be divided into group (G) and individual (S). According to the combination of observation type and record type, they can be divided into group visual observation (VG), group measurement (MG), individual visual observation (VS) and individual measurement (MS).

Expression state: In the DUS test guide or standard for plant varieties, the expression range of each test trait is divided into a series of expression states. To facilitate the definition of traits and specification descriptions, each expression state is assigned a corresponding numeric code to facilitate test data recording, processing and variety description.

Preferably, the making of a unified table includes making a unified parameter table, and the parameter table includes at least fields for the following parameters: code, standard value, expression status, standard variety, trait number, trait name, and numerical type; preferably, the parameter table also includes one or more parameter fields selected from the following: expression type, observation type, observation time, quantity unit, grading value, maximum value, minimum value, code index, grading value index, grouping, weight, threshold and photo.

Among them, the trait number, code, expression status, standard variety, trait name, expression type, observation type, observation time, quantity unit, numerical type, maximum value and minimum value parameters can be preset according to the DUS test guide.

The standard value is determined according to the DUS test guide or other methods.

The actual measured values of the standard varieties are the data measured in this test.

The code index is to set an identification code for each code, which can be set to be composed of "characteristic number*10000+code", so as to facilitate the subsequent extraction of corresponding information through the characteristic number and code.

The grading value index is another identification code set for each code, which can be set to be composed of "trait number *10000+grading value" to facilitate the interval method to convert the original value into a code.

The grading value is used to set the grading interval of the code corresponding to the original value, and is the minimum value of the grading interval corresponding to each code.

The code index, classification value index and classification value are automatically calculated from the standard value and the measured value according to the preset formula.

Grouping: used in the grouping procedure. The basis for grouping is the grouping trait, which is documented in the DUS test guide. However, considering the actual use effect, quality traits and/or pseudo-quality traits with discrete expression states, easy to distinguish, and accurately observable can be re-selected as grouping traits.

Weights are used in calculating the value of varieties and are set based on experience. For example, if the growth period is important, the weight is 3, and if the trait is not important, the weight is 1. This is applicable in giving scores in variety value evaluation and evaluating the significance of variety planting from a technical perspective.

The threshold is used in the threshold method to screen similar varieties. It is set empirically based on the variation of data over the years. For example, the quality trait is set to 0, the pseudo-quality trait is set to 1, and the quantitative trait is set to 2. If the preset threshold is found to be inappropriate after the increase of samples or technical improvements, appropriate adjustments can be made.

Photos: For each type of object photographed, specific characters (such as digital numbers) are manually assigned in advance and the photos are named with the characters. For example, corn photos are divided into five types: seedlings, plants, tassels, filaments, and ears. They are numbered 1, 2, 3, 4, and 5 respectively. The photos are named with these numbers, for example 1.jpg, which represents a seedling photo.

To facilitate analysis and processing, the storage management of photos can be optimized and folders at all levels can be established in a unified way, such as photos\corn\2019\variety name, and the photos can be stored in folders with the corresponding variety names.

According to the corresponding relationship, the characters used to name the photo (such as serial numbers) are input into the photo field of the corresponding trait and linked with the corresponding photo.

In order to connect with other systems, there may be situations where the names of photos or folders need to be uniformly modified. In this case, a table is created according to the field format of the old name, file type, file address, and new name, and the new name of the photo is entered. The file or folder batch renaming can be completed by linking the folder and/or photo through the program.

As an example, see Table 1.

Table 1

Regarding the settings of data format (such as data type), value range and optional value unit, you can use the table or database built-in function to set the data type. Taking EXCEL cells as an example, the data types include arbitrary number, integer, decimal, sequence, date, time, text length, etc. For code-type visual data, you can select sequence and fill in the allowed code value in the data source; for continuous measurement data, you can select decimal and fill in the allowed minimum and maximum values in the data source; for discrete measurement data, you can select integer and fill in the allowed minimum and maximum values in the data source; for date-type data, you can select date and fill in the start date and end date in the data source; for colorimetric card data, you can select text length and fill in the minimum length 4 and maximum length 5 in the data source, etc. Taking corn plant height as an example, the value range can be set to 30-500, and the value unit can be set as needed, such as cm.

Preferably, data is collected using a table in a unified format; for example, data is collected using a horizontal data table or a vertical data table, preferably, data is collected using a horizontal data table; the format of the horizontal data table is: arranged horizontally according to the fields of the test, variety, test, and trait number; when multiple single-plant sample values are measured for the same trait, the same trait number is repeated horizontally; the same variety under the same test is only listed once, and no duplication occurs.

The format of the vertical data table is as follows: the fields of the test, variety, test, trait, and sample number of each individual plant of the same trait are arranged horizontally, and the trait number is used as data in a vertical arrangement.

In each table, the varieties marked with "Yes" under the field to be tested indicate the varieties that need to be tested and evaluated and for which an analysis report needs to be issued. Other varieties are indicated by "No", such as standard varieties and similar varieties, which are not tested and evaluated varieties and do not require an analysis report.

Preferably, in the above-mentioned plant variety DUS test method, the test of the validity of the test data also includes using the BoxPlot method (box plot method) and/or the 3σ method (three times standard deviation method) to test the MS data of multiple samples collected in the test; if an abnormal value appears, it is automatically marked. If it is still an abnormal value after the two methods are tested, it is necessary to manually check the original records or field samples. If it is an input error, it is directly corrected. In the case of objective facts, if there are only a few (for example, less than two) abnormal values that cannot explain the reason, the average of the previous and next values provided by the program can be corrected (other calculation methods such as the overall mean value and the maximum likelihood method are not suitable for the correction of DUS test data. The test results of the plants near the plant are used to estimate the missing values, which can better reflect the actual situation of the test). If there are many abnormal values, they will not be processed, and the relative variance method or COYU method can be considered to test the consistency. Abnormal values may also be caused by the environment or sampling method, such as uneven soil fertility, inconsistent environment, or sampling without excluding marginal plants. This requires optimizing the experimental design and sampling methods, and expanding the sampling quantity when necessary.

For traits measured by repeated sampling, the use of BoxPlot or the 3σ method can more accurately test outliers. In the former, extreme values are not included in the calculation, while in the latter, extreme values are included in the calculation. The two methods are complementary.

Preferably, a vertical data table is used for BoxPlot method and/or 3σ method test; when data is collected using a horizontal data table, it can be converted into a vertical data table through a designed program. The above two test methods can be performed in the same data format, and various abnormal values can be marked with different colors.

Preferably, the step of collecting and processing the test data in a unified format further includes converting the raw data of the trait into codes; the step includes:

For each visual trait VG and VS raw data, directly assign the corresponding code and conduct the plant variety DUS test;

For each measured trait MG and MS raw data, frequency distribution analysis is performed, including: selecting the trait raw data of any one test or the trait raw data of any multiple tests for LSD analysis (any one or multiple test data are acceptable, and the test with as many varieties as possible can be calculated once, without calculating for each test), and obtaining the LSD _0.05 value of the measured trait; finding the average value of all measured trait raw data of all varieties; taking the average value of all measured trait raw data of all varieties as the central standard value, and 2 times LSD _0.05 as the level difference, setting the standard values of each level; taking the standard value of each level as the center, extending the interval formed by 1/2 level difference to both sides, and setting it as the classification interval of each code; the minimum value of each interval is taken as the classification value; and counting the number and percentage of varieties in each classification interval;

(2) Determine the standard value, grading value and interval code, including: according to the statistical results, determine whether the total interval coverage is less than 3 levels or greater than 9 levels; eliminate the traits less than 3 levels and are not suitable for DUS analysis of plant varieties; for traits greater than 9 levels, increase the multiple of LSD _0.05 , and adjust the grading by referring to whether the percentage of each grading interval is uniform, so that the grading range is within 9 levels; for traits between 3 and 9 levels, leave 1-2 levels at both ends to be used when new varieties appear in the future; if the minimum value of the minimum grading interval is less than zero, set the minimum value of the minimum grading interval to 0, and then grade from small to large according to the multiple of LSD _0.05 determined previously, and re-determine the standard value; preferably, the obtained standard value is accurate to an appropriate number of decimal places (for example, 0 decimal places, i.e., rounding);

Select a plant variety whose measured value of the measured trait is at or near the standard value of each level as the standard variety of the corresponding classification interval. When the error is large, appropriately translate the standard value to make the measured value of the standard variety close to its corresponding standard value, thus eventually forming a set of relatively fixed standard values and standard varieties. The selection of standard varieties should take into account the adaptability of the variety, the availability of propagation materials and the representativeness of the expression state.

In different test trials, the standard value usually remains unchanged and can be verified in subsequent steps. When an unsuitable situation occurs, such as a new expression state, it can be readjusted and confirmed. The standard variety is usually kept unchanged and planted repeatedly in each trial. The measured value of the standard variety and the grade value determined by it will change from trial to trial.

Based on the given grading intervals and codes, the raw data of the measured traits MG and MS in the experiment are coded to obtain interval codes, which can be directly used for DUS analysis of plant varieties, or the interval codes can be further optimized and used for DUS analysis of plant varieties.

Preferably, the above-mentioned frequency distribution analysis is performed in a horizontal cross-experiment data table. When performing frequency distribution analysis on the MS trait data of one experiment, the test items, varieties, trait numbers and their original values are directly extracted from the horizontal row data table to the horizontal cross-experiment data table; when performing frequency distribution analysis on the MS trait data of two or more experiments, the test items, varieties, and trait numbers are extracted from the horizontal row data table, and the experimental average values of each trait are calculated and transferred to the horizontal cross-experiment data table. The format of the horizontal cross-experiment data table is: the test items, varieties, and average values of different experiments of the same trait or the original values of different plants are arranged in consecutive horizontal rows.

Preferably, in the above-mentioned plant variety DUS test method, for MG and MS traits, when the set of standard values adopted is not obtained from the current test data, it also includes a step of verifying whether the performance of the standard variety in this test is consistent with that in the test in which the set of standard values is obtained, and the step includes: comparing the actual measured value of the standard variety in this test with the corresponding standard value, dividing the absolute value of the difference between the two by the standard value, and if the value is greater than 10%, it is confirmed as an abnormal value and marked (for example, marked with a specific color); in the case of abnormal values, it is manually determined whether it is an acceptable abnormal situation. If multiple standard varieties of a certain trait show similar changes due to some factors, it is determined to be an acceptable abnormality; if the standard variety with the abnormal situation of the trait is inconsistent with the changes of other standard varieties, it is necessary to eliminate the actual measured value of the standard variety.

Preferably, in the step of checking whether the performance of the standard variety in this test is consistent with that in the test for obtaining the set of standard values, the measured value, standard deviation and sample number of the trait in this test are calculated in the vertically processed data table, and the measured value of the standard variety is extracted into the parameter table for testing; the vertically processed data table can be converted from the vertical data table, and the format of the vertically processed data table is: test, variety, test, and trait are arranged horizontally, and multiple sample values of the same MS trait in the vertical data table are processed into average value, standard deviation and sample number, and fields for interval code, known code, regression code, optimization code and expression status are reserved, and these fields are arranged horizontally.

Preferably, for MG and MS traits, when the set of standard values adopted is not obtained from the test data of this time, the step of correcting the grading range using standard varieties is also included, and the step includes: the first grading value is zero, and the second and subsequent grading values are: the sum of the values obtained by (measured value-standard value)/standard value of each standard variety divided by the number of standard varieties, plus 1/2 of the sum of the standard value corresponding to the current code and the standard value corresponding to the previous code.

Preferably, in the above-mentioned plant variety DUS testing method, for quantitative traits, it also includes establishing a linear regression function using the known codes of known varieties (including standard varieties and/or similar varieties) and the corresponding average values of the known varieties in this experiment, and substituting the average value of the original data of the variety to be tested in this experiment into the regression function to calculate its regression code; and analyzing the interval code, known code, and regression code, and selecting the mode code, median code or the average code of the three as the code data for further optimization.

Preferably, in the above-mentioned plant variety DUS test method, when there are at least two codes among the interval code, the known code and the regression code, it also includes a verification step using the code range, and the verification step includes: calculating the range by subtracting the minimum value from the maximum value among these codes, and marking different sizes of ranges differently; for example, a code difference of 1 is displayed in yellow, a code difference of 2 is displayed in orange, a code difference of 3 is displayed in red, and a code difference of 4 or more is displayed in purple. For the marked code data, the original data is manually checked, or photos are retrieved for confirmation, and the optimized code data is manually modified as needed.

Preferably, in the above-mentioned plant variety DUS test method, the optimized codes of multiple tests are put together for comparison, the code range obtained by using the maximum value-minimum value is used to test the code, different markings are made according to the range size, such as displaying in different specific colors, and the marked code data is manually checked for original data, or photos are retrieved for confirmation, and the code is manually modified as needed to obtain a comprehensive code across the tests;

Preferably, the inspection of the code using the code range is carried out in a vertical cross-test data table, and the format of the vertical cross-test data table is: the test, variety, trait, the average value, standard deviation, number of samples, code, and expression status in each test are displayed horizontally side by side, and the average value, standard deviation, number of samples, and average value of the optimized code of all tests are calculated and displayed horizontally side by side, wherein the average value of the optimized code is rounded, that is, the value after the decimal point is directly removed or rounded off; the code range of different tests is calculated, and different identifications are performed according to the size of the range, for example, displaying in different specific colors, such as yellow (difference 1), orange (difference 2), red (difference 3), and purple (difference 4 or more); for the colored code data, that is, the code that is different between each test, the original data is manually checked, or photos are retrieved for confirmation, and the code is manually modified as needed, and the confirmed code is used as the cross-test comprehensive code.

Preferably, in the above-mentioned plant variety DUS test method, the step of transferring the converted code into a variety library is also included; preferably, if the variety or corresponding trait already exists in the variety library, the result is overwritten; if not, the variety is added to the last row of the variety library, or the trait is added to the last column, and the variety library is updated;

Preferably, after completing two or more tests, the photos of the two or more tests are compared one by one to determine whether there are differences between the two or more tests. If there are differences, check the reasons. If there are no differences, select a set of standard photos and store them in a preset folder, for example, in the DUS\Corn\Standard Photos\Variety Name folder.

Preferably, the process further includes the step of confirming or correcting the trait code using photos; preferably, after arranging a certain trait code in ascending order, extracting photos corresponding to the trait and visually comparing them in sequence; preferably, the process is performed in a horizontal data table or variety library, and the mouse is placed on a certain code-type trait column. The program automatically sorts the codes by size, and extracts photos of each variety corresponding to the trait in batches using the characters (preferably digital numbers) named in the photos, places them in the corresponding positions in the next column, and checks the photos in sequence to manually confirm whether any of the codes are incorrect, ultimately ensuring that the codes and photos to be reported are consistent.

The method of the present invention optimizes the DUS test design, realizes the joint correction of multiple test data, and increases the objectivity of the DUS test analysis results. At the same time, the method of the present invention can efficiently realize the test data and statistical hypothesis validity test and DUS data analysis with the help of EXCEL program, so that the DUS data analysis work that originally took 2 months to complete can be reduced to 1 day.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 shows a vertically arranged processed data table showing linked photos;

Figure 2 shows a vertical cross-experiment data table showing linked photos;

Figure 3 shows an example of the comparison interface of the two experiments;

Figure 4 shows an example of the interface for photo confirmation in the variety library;

FIG5 shows an example of an interface for comparing photos of a variety to be tested with similar varieties in the variety library;

DETAILED DESCRIPTION

The specific embodiments of the present invention are described in detail below. It should be understood that the specific embodiments described herein are only used to illustrate the present invention by way of example and are not used to limit the present invention.

The experimental crop is corn, with 149 varieties, including 147 varieties and 2 standard varieties. It is planted once in the first year and once in the second year. Each plot is 5 meters long and 2.4 meters wide, with four rows planted, and the row spacing is 30cm×60cm. Double seeds are sown, and two leaves and one heart are planted. At least 80 seedlings are left in each plot, and two replicates are set. The field management measures are the same as those in field production. When collecting data, only one value is collected for VG and MG traits, and 20 single plant values are collected for VS and MS traits. Photos of seedlings, plants, tassels, filaments, and ears are taken, and the samples are all from the same plot. Five photos of seedlings, plants, tassels, filaments, and ears are taken during the whole growth period.

1. Set the parameter table for corn according to the DUS test guide

Due to the large number of traits, only three traits are taken as an example below. The parameter table settings are shown in Table 2:

Table 2

In Table 2, parameters such as trait number, code, expression status, standard variety, trait name, expression type, observation type, observation time, quantity unit, numerical type, maximum value, minimum value, etc. can be preset according to the DUS test guide, and the standard value is determined according to the above method; the threshold value and weight are set according to experience; the code index, classification value index, and classification value are automatically calculated from the standard value and the measured value according to the preset formula.

2. Horizontal data table records original data

The record format is shown in Table 3. Due to the large number of varieties and quantities, only the first 16 traits and the first 14 varieties are listed.

Table 3

According to the unified horizontal data format, horizontal data collection is performed to create a horizontal data table: the fields of test, variety, test, and trait number are arranged horizontally. When multiple single plant sample values are measured for the same trait, the same trait number is repeated horizontally. For example, in Table 3, 20 single plant samples were measured for the trait numbered 16, and the trait number 16 is arranged horizontally continuously.

In Table 3, the varieties marked with "yes" in the field to be tested indicate the varieties that need to be tested and evaluated and for which an analysis report needs to be issued. Other varieties are indicated by "no", such as standard varieties and similar varieties, which are not tested and evaluated varieties and do not require an analysis report.

The same variety under the same test can only be listed once and no duplication is allowed.

3. Data verification

(1) Data format (e.g., data type), value range, and/or optional value unit matching check

Verify the horizontal data based on the data format such as data type, value range such as minimum value, maximum value, and value units such as day and cm in the parameter table in Part 1.

In the test results, outliers are marked with a specific color, such as red. For example, based on the settings in the parameter table in Part 1, the value range of trait 2 should fall within the range of 1-5, and the value 6 entered during data collection is an outlier; the value range of trait 16 should be 10-70, and the value 4 entered during data collection is an outlier. These outliers are automatically displayed in red through the design program.

(2) Use BoxPlot method and 3σ method to test the validity of test data

Convert the horizontal data format into the vertical data format and make a vertical data table. Use the BoxPlot method and the 3σ method to test the validity of the test data.

The converted vertical data table is shown in Table 4. The fields of the test, variety, test, trait, and sample number of each individual plant of the same trait are arranged horizontally, while the trait number is used as data in vertical arrangement.

Table 4

The same vertical data table can be used to perform both calculations, with various outliers identified by different colors.

The calculation results of the BoxPlot method are shown in Table 5 (yellow represents 1.5 times the center distance, and red represents 3 times the center distance).

Table 5

The calculation results of the 3σ method are shown in Table 6 (yellow is 2 times the standard deviation, red is 3 times the standard deviation).

Table 6

If the value is still an outlier after the two methods are tested, the original record or field sample needs to be manually checked. If it is an input error, it should be corrected directly. In the case of objective facts, if there are only a few (for example, only two or less) outliers that cannot be explained, the program will provide the average of the before and after values. If there are many outliers, they will not be processed and the relative variance method or COYU method will be used to test the consistency.

4. Frequency distribution analysis

(1) When performing frequency distribution analysis on MS trait data from one test, the test items, varieties, trait numbers and their original values can be directly extracted from the horizontal data table into the horizontal cross-test data table; (2) When performing frequency distribution analysis on MS trait data from two tests, the test items, varieties, trait numbers can be extracted from the horizontal data table, and the experimental average values of each trait can be calculated and transferred into the horizontal cross-test data table. The format of the horizontal cross-test data table is: the test items, varieties, and the average values of different tests of the same trait or the original values of different plants are arranged in a row.

The horizontal cross test data table is shown in Table 7.

Table 7

Frequency distribution analysis and determination of standard values and grading values

An example of the calculation results of the frequency distribution of quantitative traits is shown in Table 8.

Table 8

Characteristics 16 17 18 19 twenty two 25.2 26.2 27.2 29.2 30.2 31.2 Overall mean 40.68 29.72 7.956 25.19 10.58 103 276.6 0.372 19.96 4.599 16.26 sum 57769 42198 11297 35771 15019 1E+05 4E+05 528 28338 6530 23088 Total variance 18.24 15.54 9.829 13.03 0.92 276.7 711 0.002 3.612 0.068 3.169 Total sum of squares 2E+06 1E+06 1E+05 9E+05 2E+05 2E+07 1E+08 199 6E+05 30125 4E+05 LSD0.05 1.63 1.488 1.079 1.649 0.424 5.11 6.855 0.018 0.723 0.094 0.846 16 Median Minimum quantity frequency Minimum average value Maximum 1 34.1 40.68 51.2 2 3 34.16 32.53 8 0.113 4 37.42 35.79 11 0.155 5 40.68 39.05 32 0.451 6 43.94 42.31 14 0.197 7 47.2 45.57 4 0.056 8 50.46 48.83 2 0.028 9 17 Median Minimum quantity frequency Minimum average value Maximum 1 23.75 29.72 37.25 2 3 23.77 22.28 5 0.07 4 26.74 25.25 16 0.225 5 29.72 28.23 25 0.352 6 32.69 31.2 19 0.268 7 35.67 34.18 5 0.07 8 38.64 37.16 1 0.014 9

For each measured trait, the average value of all original data of all varieties is used as the central standard value, and 2 times LSD _0.05 is used as the grade difference to set the standard value of each level. The interval formed by extending 1/2 grade difference to both sides with the standard value of each level as the center is set as the grading interval of each code; the minimum value of each interval is used as the grading value; the grading value and grading interval of each level are determined, and the number and percentage of varieties in each grading interval are counted. According to the statistical results, it is determined whether the total interval coverage is less than 3 levels or greater than 9 levels. Traits less than 3 levels are not suitable for DUS testing and are eliminated; for traits greater than 9 levels, the multiple of LSD _0.05 is increased, and the grading is adjusted by referring to whether the percentage of each level interval is uniform, so that the grading range is within 9 levels; for traits between 3 and 9 levels, 1-2 levels can be left blank at both ends to be applicable when new varieties appear in the future; thus, a more reasonable grading is obtained. If the minimum value of the minimum grading interval is less than zero, the minimum value of the minimum grading interval is set to 0, and then the grading is carried out from small to large according to the LSD _0.05 determined previously, and the standard value is re-determined, and the standard values of each level are rounded. Then, according to the performance of the standard variety in the test, it is determined whether the code position of the standard variety is appropriate. If it is not appropriate, the standard variety is moved to a suitable code position, and according to the measured value of the standard variety, the standard value is shifted as a whole, so that the measured value of the standard variety is close to its corresponding standard value. In this way, the standard variety is determined and a set of relatively fixed standard values is formed. As shown in Table 9.

Table 9

5. Convert vertical data format to vertical processing data format, create vertical processing data table, and perform code data inspection and optimization

The vertical data format is converted into a vertical processing data format, that is, the test, variety, test, and trait are arranged horizontally, and multiple sample values of the same MS trait in the vertical data table are processed into average value, standard deviation, and number of samples, and the interval code, known code, regression code, optimization code, and expression status fields are reserved and arranged horizontally. The processed table format is shown in Table 10.

Table 10

For MG and MS traits, if the set of standard values used is not obtained from the current test data (for example, the second year test), check whether the performance of the standard variety in this test is consistent with that in the test where the set of standard values was obtained.

In the vertical processing data table, calculate the average value (i.e. measured value), standard deviation, and number of samples of the trait in this test. Extract the measured value of the standard variety into the parameter table. The absolute value of the difference between the measured value of the standard variety and the standard value is divided by the standard value. If the value is greater than 10%, it is confirmed that the difference is too large, and a specific color (such as red) is used to display the prompt as an abnormal value. In the case of abnormal values, manual judgment is made as to whether it is an acceptable abnormal situation (in principle, quality traits are not allowed to have abnormalities. If they do occur, manual inspection of the cause is required and corrections are made). If multiple standard varieties of a certain trait show similar changes due to some factors, it is an acceptable abnormality and the data is retained. If the standard variety with abnormal conditions for the trait is inconsistent with the changes of other standard varieties, the measured value of the standard variety needs to be eliminated.

For MG and MS traits, when the standard values used are not obtained from the test data, the standard varieties are used to correct the grading range. The first grading value is zero, and the second and following values are: the sum of the values obtained by (measured value-standard value)/standard value of each standard variety divided by the number of standard products, plus 1/2 of the sum of the standard value corresponding to this code and the standard value corresponding to the previous code.

The grading value, grading value index and grading interval corresponding to each code are calculated according to the standard value and the measured value, and the code of each trait of each variety is calculated to obtain the interval code.

For quantitative traits, if there are known varieties with data in the variety library in this test, the corresponding codes of the known varieties in the variety library are mentioned in the known codes in the vertical processing data table. A linear regression relationship is established between the average values of these known varieties in this test and the corresponding known codes, and the average values of the original data of other varieties in this test are substituted into the linear regression relationship to calculate the regression codes of all varieties. The interval codes, known codes, and regression codes are analyzed, and the mode code, the median code, or the average code of the three is selected as the code data for further optimization.

When there are at least two codes among the interval code, known code, and regression code, the code range is used for inspection. The range is calculated by subtracting the minimum value from the maximum value of these codes, and a specific color is displayed according to the range size. For example, yellow is displayed when the difference is 1 code, orange is displayed when the difference is 2 codes, red is displayed when the difference is 3 codes, and purple is displayed when the difference is 4 codes. For the color code data, the original data can be manually checked, or photos can be retrieved for confirmation. The code can be manually modified as needed. By designing a program, photos can be quickly retrieved for confirmation. Put the mouse on the code that needs to be retrieved, link the photo through the program, and display it in the blank position of the same line of the code. As shown in Figure 1.

6. Convert the vertical processing data format to the vertical cross-test data format and determine the cross-test comprehensive code

The vertical processing data format of the two-year test is converted into a vertical cross-test data format, and a vertical cross-test data table is prepared, as shown in the following table. The format of the vertical cross-year data table is: the test, variety, trait, average value, standard deviation, number of samples, code, and expression status in each year's test are displayed horizontally and side by side, and the average value, standard deviation, number of samples, and average value of the optimized code of the two-year test are calculated and displayed horizontally and side by side. Among them, the average value of the optimized code is rounded (the value after the decimal point is directly removed or rounded). In addition, the range of each year's code is calculated, and different specific colors are displayed according to the range size, such as yellow (difference 1), orange (difference 2), red (difference 3), and purple (difference 4 or more). For the color code data (codes with differences between years), the original data is manually checked, or photos are retrieved for confirmation, and the code can be manually modified as needed. The confirmed code is used as a cross-test comprehensive code. By designing a program, it is possible to quickly retrieve photos for confirmation. Put the mouse on the code that needs to be retrieved, link the photo through the program, and display it in the blank position of the same line of the code. As shown in Figure 2.

7. Transfer cross-trial integrated codes to the variety library

The comprehensive code data in the cross-experimental data table is transferred to the variety library, as shown in Table 11. For varieties or traits already in the variety library, the code data is directly overwritten; for varieties or traits not in the variety library, a field is added to the last row (variety) or last column (trait) of the data, and the comprehensive code data is imported. In this way, the data in the variety library is updated and accumulated year by year.

Table 11

To be tested variety 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 twenty one twenty two twenty three twenty four 25 26 yes DK145 2 2 5 5 4 2 4 2 2 2 4 4 4 2 2 6 6 2 6 2 2 4 2 2 8 7 yes DK817 6 4 4 5 4 2 4 3 2 2 2 5 4 4 1 6 8 2 6 2 2 5 2 2 8 8 yes FH218 4 4 5 5 4 2 4 2 2 2 2 4 2 2 2 6 6 2 5 2 4 5 2 2 8 7 yes MC598 2 4 4 5 4 2 4 3 2 2 4 4 4 1 1 5 5 2 4 2 2 5 2 2 6 6 yes MC858 4 4 4 5 4 3 4 2 2 2 5 4 4 1 3 6 6 2 6 2 3 5 2 2 7 6 yes ND688 2 4 4 5 4 1 4 3 2 4 4 4 4 2 2 5 5 2 6 2 3 4 2 2 8 6 yes Beinong 486 5 4 5 5 4 2 4 4 2 4 4 6 4 1 2 6 6 2 6 2 2 5 2 2 8 6 yes Beinong 851 4 4 5 6 6 2 4 4 2 2 2 4 4 1 4 5 5 4 6 2 2 5 2 2 8 7 yes Beinong 861 2 4 5 6 6 2 4 3 2 2 3 4 4 2 2 6 5 2 6 2 2 5 2 2 10 8 yes Beinong Silage 36 6 4 5 6 6 2 4 2 2 4 2 4 4 1 2 5 3 2 5 2 3 5 2 2 11 10 yes Changyu88 2 2 5 5 4 3 4 2 6 2 2 4 4 2 2 5 4 2 6 2 3 5 2 2 8 6 yes Datang 121 2 4 4 4 4 2 4 2 2 2 2 5 4 1 1 5 5 2 4 2 2 4 2 2 8 6 no Decca 517 2 3 4 4 4 2 4 2 2 2 4 4 4 1 1 5 4 4 5 2 3 4 2 2 8 6 yes Heyu 501 3 4 4 4 4 2 4 3 2 2 2 4 4 1 1 6 6 2 6 2 3 4 2 2 6 6

8. Standard photo arrangement

After completing the two-year experiment, compare the photos of the two years. Determine whether there are any differences between the two years. If there are differences, check the reasons. If there are no differences, select a set of standard photos and store them in the DUS\Corn\Standard Photos\Variety Name folder. An example of the comparison interface is shown in Figure 3.

9. After the codes are sorted, batch retrieve photos to confirm the codes

In the horizontal data table or variety library, click the mouse on a certain code-type trait column, the program will automatically sort by code and size, and preset the photo type number corresponding to the trait under the trait photo field in the parameter table of Part 1, batch extract the photos of each variety corresponding to the trait, put them in the corresponding position of the next column, check the photos in order, and manually confirm whether there is a wrong code. Finally, ensure that the code and photo to be reported are consistent. The interface example of photo confirmation in the variety library is shown in Figure 4.

10. Specificity analysis

(1) Screening similar varieties in the variety library

In the final variety library, all varieties are first grouped and sorted according to the grouping information set in the parameter table. Varieties can be quickly and preliminarily divided into several groups. If there is only one variety in a group, it does not need to be compared with other varieties.

Then, the varieties were analyzed for their approximation degree using the accumulation method of the number of traits with differences, the accumulation method of the number of traits with differences greater than the threshold, the correlation coefficient method, or the minimum distance method, and the general warning and special warning value ranges were set respectively, as shown in Table 12.

Table 12

Taking the correlation coefficient as an example, in the result area, the horizontal row is the varieties to be tested, and the vertical row is all varieties. Correlation coefficients greater than 90% are displayed in yellow, and those greater than 95% are displayed in red.

(2) Further confirmation using variety photo comparison

Delete all data under the field to be tested in the variety library table, fill in "yes" before the variety to be tested that needs further comparison, and fill in "no" before the similar variety analyzed in step (1). Based on the input information, the program sequentially retrieves photos of the variety to be tested and similar varieties, displays them side by side, and manually quickly checks and confirms whether there are any differences. An example of the interface for comparing the photos of the variety to be tested and similar varieties is shown in Figure 5.

(3) Retrieving original data

If there is no difference between the photos of the similar variety and the variety to be tested, further retrieve the average values and codes (MS traits also include standard deviations) of all measured traits of the two varieties from the vertical processing data table or the vertical cross-test data table and place them in the data comparison table for side-by-side comparison, as shown in Table 13, to see if there are data with large differences.

Table 13

(4) For the MG and MS traits, the original data of a single test or the summary data of two tests of the two varieties were retrieved for T test or COYD analysis, as shown in Tables 14 and 15.

Table 14

Table 15

(5) For VS traits, in the vertical row and column data table, the Pearson chi-square test method was used to analyze the specificity of the tested variety and similar varieties.

The original data can be converted from the vertical data format to the vertical row and column data format through statistics. The format of the vertical row and column data table is: the test, variety, test, trait, code field is arranged horizontally, and the data under each code field is the number of times the code appears in the population, as shown in Table 16. You can also directly collect field data in the vertical row and column data format.

Table 16

To be tested variety test Characteristics 1 2 3 yes C01 2018 1 34 6 6 no R01 2018 1 12 twenty three 9 no R02 2018 1 6 20 19 no R03 2018 1 1 18 9 no R04 2018 1 7 twenty two 15 yes C02 2018 1 9 3 34 no R05 2018 1 4 8 34 no R06 2018 1 1 11 34

The results of the Pearson chi-square test are shown in Table 17.

Table 17

To be tested variety test Characteristics 1 2 3 C01 result C02 result yes C01 2018 1 34 6 6 1 2E-08 D no R01 2018 1 12 twenty three 9 3E-05 D 3E-07 D no R02 2018 1 6 20 19 4E-08 D 0.0002 D no R03 2018 1 1 18 9 2E-08 D 5E-07 D no R04 2018 1 7 twenty two 15 2E-07 D 2E-05 D yes C02 2018 1 9 3 34 2 2E-08 D no R05 2018 1 4 8 34 3E-10 D 0.1227 ND no R06 2018 1 1 11 34 5E-12 D 0.0041 D

(6) When there are only two expression states in (5), the more accurate Fisher's exact test method can be used to analyze the specificity.

The calculation results of Fisher's exact test are shown in Table 18.

Table 18

To be tested variety test Characteristics 1 2 C01 result C02 result yes C01 2018 1 34 6 0.2295 ND no R01 2018 1 12 twenty three 6E-06 D 0.0146 ND no R02 2018 1 6 20 5E-07 D 0.0033 D no R03 2018 1 1 18 3E-09 D 9E-05 D no R04 2018 1 7 twenty two 4E-07 D 0.0033 D yes C02 2018 1 9 3 0.2295 ND no R05 2018 1 4 8 0.0011 D 0.0436 ND no R06 2018 1 1 11 2E-06 D 0.0013 D

11. Discrete data consistency analysis

The consistency analysis of discrete data adopts the UPOV heterotypic strain method. This method transforms the display interface to facilitate data recording and batch calculation. The interface is shown in Table 19.

Table 19

To be tested variety test Characteristics Total number of plants Number of heterotypic strains Heterotypic strain method consistency no R01 2017 1 20 3 0.978991644 NU no R02 2017 1 30 3 0.996682291 NU no R03 2017 1 40 3 0.992502637 NU no R04 2017 1 50 3 0.810798075 U no R05 2017 1 60 3 0.73146611 U no R06 2017 1 70 3 0.649236912 U no R07 2017 1 80 6 0.966693618 NU no R08 2017 1 90 6 0.946079612 U no R09 2017 1 100 6 0.919162871 U no R10 2017 1 110 6 0.886011935 U no R11 2017 1 120 6 0.847081858 U yes C01 2017 1 130 6 0.803137839 U yes C02 2017 1 140 9 0.974122695 NU yes C03 2017 1 150 9 0.962190432 NU yes C04 2017 1 160 9 0.946973552 U yes C05 2017 1 170 9 0.928236031 U yes C06 2017 1 180 9 0.905863816 U yes C07 2017 1 190 9 0.879872217 U

The P value and results are calculated based on three values: population size, number of heterotypic strains, and population standard.

12. Continuous data consistency analysis-relative variance method;

The one-year data can be analyzed for consistency using the relative variance method. The data collection and analysis interface is shown in Table 20.

Table 20

To be tested variety test Characteristics Total number of plants Standard Deviation Relative variance Allow relative variance consistency no R01 2017 1 20 1.5 1.012269939 1.878311735 U no R02 2017 1 20 2.2 1.484662577 1.878311735 U no R03 2017 1 20 2.3 1.552147239 1.878311735 U no R04 2017 1 20 1.4 0.944785276 1.878311735 U no R05 2017 1 20 0.5 0.337423313 1.878311735 U no R06 2017 1 20 1.4 0.944785276 1.878311735 U no R07 2017 1 20 1.4 0.944785276 1.878311735 U no R08 2017 1 20 1.4 0.944785276 1.878311735 U no R09 2017 1 20 1.4 0.944785276 1.878311735 U no R10 2017 1 20 1.4 0.944785276 1.878311735 U no R11 2017 1 20 1.4 0.944785276 1.878311735 U yes C01 2017 1 20 1.7 1.147239264 1.878311735 U yes C02 2017 1 20 4.1 2.766871166 1.878311735 NU yes C03 2017 1 20 1.53 1.032515337 1.878311735 U yes C04 2017 1 20 1.47 0.99202454 1.878311735 U yes C05 2017 1 20 1.41 0.951533742 1.878311735 U yes C06 2017 1 20 3 2.024539877 1.878311735 NU yes C07 2017 1 20 1.29 0.870552147 1.878311735 U

13. Continuous data consistency analysis-COYU method

The two-year data adopts the COYU method. The data can be obtained from the horizontal cross-test data table. The analysis interface is shown in Table 21.

Table 21

Claims (6)

1. A plant variety DUS testing method comprising performing a plant planting test, collecting and processing test data in a unified format, and performing DUS analysis, the variety types in the plant planting test comprising a variety to be tested, a standard variety, and optionally an approximate variety; it is characterized in that the method comprises the steps of,

the step of collecting and processing test data in a unified format includes the step of converting character raw data into codes; the method comprises the following steps:

directly giving corresponding codes to the original data of each visual trait VG and VS, and carrying out plant variety DUS test;

for each measurement trait MG and MS raw data, a frequency distribution analysis was performed comprising: selecting character original data of any one test or character original data of any plurality of tests to carry out LSD analysis to obtain LSD for measuring characters _0.05 A value; calculating the average value of all the original data of the measured characters of all varieties; taking the average value of the raw data of the measured properties of all varieties as a central standard value and taking 2 times of LSD _0.05 Setting standard values of all levels for the level difference; taking standard values of all levels as centers, and setting a section formed by extending 1/2 level difference to two sides as a grading section of each code; the minimum value of each interval is used as a grading value; counting the number and percentage of varieties in each grading interval;

determining standard values, hierarchical values, and interval codes, comprising: judging whether the total interval coverage is smaller than 3 levels or larger than 9 levels according to the statistical result; less than 3 grades of characters are removed, and the method is not suitable for DUS analysis of plant varieties; traits greater than grade 9, modulating LSD _0.05 And (2) is a multiple ofAdjusting the classification according to whether the percentages of the classification intervals are uniform or not so that the classification range is within 9 stages; traits between 3 and 9 levels, which can be used when 1-2 levels are left at each end for future appearance of new varieties; if the minimum value of the minimum classification interval is smaller than zero, the minimum value of the minimum classification interval is set to 0, and the LSD is further processed according to the above determined multiple _0.05 Grading from small to large, and re-determining a standard value;

selecting a plant variety with the measured property actual measurement value at or near each level of standard value as a standard variety of a corresponding grading interval, and properly translating the standard value when the error is large to enable the standard variety actual measurement value to be close to the corresponding standard value, thereby finally forming a set of relatively fixed standard value and standard variety;

Coding the original data of the measurement characters MG and MS in the test based on the grading interval and the code to obtain an interval code, wherein the interval code can be directly used for plant variety DUS analysis or used for plant variety DUS analysis after the interval code is further optimized;

the collecting and processing test data in a unified format further comprises: preparing a unified form, setting a unified data format, a numerical range and optional numerical units, and checking validity of test data and statistical assumptions;

the checking of the validity of the test data comprises: performing data format, numerical range and/or optional numerical unit matching check, wherein the check is performed automatically on data during or after input by a program, if the input data does not accord with the set data format, numerical range and/or optional numerical unit, an abnormal value is indicated, if the abnormal value is found, the abnormal value is automatically identified, and an original record or a field sample is manually checked; if the input error is included, directly correcting; if the abnormal data belongs to objective facts, the abnormal data is reserved, and the following steps are continued;

The step of making the unified table comprises making a unified parameter table, wherein the parameter table at least comprises the following fields of parameters: code, standard value, expression state, standard variety, character number, character name and numerical type; the parameter table also includes one or more parameter fields selected from the following: expression type, observation time, number unit, rating value, maximum value, minimum value, code index, rating value index, group, weight, threshold value, and photograph;

the method comprises the steps of presetting character numbers, codes, expression states, standard varieties, character names, expression types, observation time, quantity units, numerical types, maxima and minima parameters according to DUS test guidelines; the standard value is determined according to the method; the threshold value and the weight are set according to experience; the code index, the grading value index and the grading value are automatically calculated by standard values and actual measurement values according to a preset formula;

wherein, a horizontal data table or a vertical data table is adopted to collect data;

the format of the horizontal data table is as follows: performing horizontal arrangement according to fields of to-be-detected, variety, test and character numbers, and continuously repeating the horizontal arrangement of the same character number when a plurality of single plant sample values are measured for the same character; the same variety under the same test is listed only once, and no repetition can occur;

The format of the vertical data table is as follows: the horizontal arrangement is carried out according to the fields of the sample numbers of each single plant of the to-be-detected, variety, test, character and the same character, and the character numbers are used as the data vertical arrangement;

the varieties marked by 'yes' under the fields to be tested in each table represent varieties which need to be tested and evaluated and need to give an analysis report, and other varieties are marked by 'no', and are not tested and evaluated varieties, and the analysis report is not needed to be given.

2. The seed plant variety DUS testing method of claim 1, wherein the testing for test data validity further comprises testing for MS data for a plurality of samples collected during the test using a BoxPlot method and/or a 3σ method; if abnormal value occurs, then automatic identification is carried out, and original record or field sample is checked manually, if the original record or field sample belongs to input error, direct correction is carried out; in the case of objective facts, if few abnormal values which cannot explain the cause occur, correction can be performed by providing an average of the values before and after the program; if the abnormal value is more, the data is not processed, and the consistency is checked by adopting a relative variance method or a COYU method.

3. The plant species DUS test method according to claim 1 or 2, further comprising comparing the codes of the plurality of tests together, checking the codes using the code range obtained by the maximum value-minimum value, performing different identifications according to the range, and manually checking the original data or retrieving photo confirmation for the identified codes, i.e., codes having differences between the tests, and manually modifying the codes as needed to obtain the integrated codes across the tests.

4. The seed plant variety DUS testing method of claim 3, wherein the code-using test is performed in a vertical cross test data table format that is: the average value, standard deviation, sample number, code and expression state in each test are displayed in parallel, the average value, standard deviation, sample number and average value of optimized code of all tests are calculated and displayed in parallel, wherein the average value of the optimized code is rounded, namely the numerical value after decimal point is directly removed or rounded; and calculating the range of different test codes, and carrying out different identifications according to the range.

5. The seed plant variety DUS testing method of claim 4, further comprising the step of transferring the obtained integrated code into a variety library; if the variety and/or the corresponding character already exist in the variety library, covering the result; if not, adding varieties in the last row of the variety library and/or adding characters in the last column, and updating the varieties.

6. The seed plant variety DUS testing method of claim 5, further comprising the step of confirming or correcting the trait code with a photograph; the operation is carried out in a horizontal data table or a variety library, the mouse is marked on a certain code character, the program automatically sorts the codes according to the size, the photos of each variety corresponding to the character are extracted in batches through the named characters of the photos, the photos are placed at the corresponding position of the next row, the photos are checked in sequence, whether the error situation of the codes is confirmed manually, and finally, the fact that the codes with reports are drawn out is ensured to be consistent with the photos.