CN111460777B - Plant variety DUS testing method - Google Patents

Abstract

A plant variety DUS testing method includes performing plant growing tests, collecting and processing test data in a unified format, and performing DUS analysis. The method optimizes the DUS test design, increases the objectivity of the DUS test result, and can efficiently realize data and statistical false validity check and DUS data analysis by means of the EXCEL program.

Description

Plant variety DUS testing method

Technical Field

The invention belongs to the technical field of DUS testing of plant varieties, and relates to a method for processing and analyzing data acquired in DUS testing, realizing DUS judgment, optimizing DUS testing guidelines and experimental design.

Background

Plant variety specificity (distictness), uniformity (Uniformity) and Stability (DUS) tests (short for DUS tests) refer to the process of evaluating the specificity, uniformity and Stability of plant varieties by planting tests or indoor analysis using corresponding test techniques and standards. DUS test is the basic technical basis of variety management in various countries, and is a necessary condition for protecting new varieties of plants, and checking or registering varieties. Has important significance for plant seed selection and breeding, promoting seed engineering construction and promoting the development of agriculture and forestry production.

Specific means that a plant variety has more than one trait that is significantly different from known varieties. Identity refers to the characteristic of a plant variety that is consistent in characteristics or behavior of the associated features or characteristics among individuals within a population, except for the natural variation that may be expected. Stability refers to the fact that a plant variety remains unchanged in its main trait after repeated propagation or at the end of a particular propagation cycle.

Internationally, the history of DUS tests is synchronized with the history of new plant variety protection regimes. 1957. The French government invited 12 Western European countries to participate in the foreign exchange held in Paris in the same year, 5 months, discussing the establishment of a special new plant variety protection system. The second outcrossing in Paris at the end of 1961 passed the International convention for New plant variety protection, and accordingly established the International New plant variety protection Association (UPOV). The convention was revised three times in 1972, 1978 and 1991, respectively, to form the most sophisticated DUS term nouns and definitions at present. The UPOV gathers member DUS test experience, organizes and drafts and gradually adopts 15 TGP technical files and 323 DUS test guidelines, and comprehensively standardizes the basic concept and principle of the DUS test, the construction and maintenance of a known variety confirmation and variety library, test experience and cooperation, development of the DUS test guidelines, test design and statistical analysis, DUS examination procedures, new plant type test guidelines, molecular technology application principles and the like.

The Chinese issues 'plant New variety protection regulations' on 3 and 20 months 1997, and DUS is listed as a necessary essential condition for variety authorization, the whole regulation content is formulated by referring to UPOV1978 text, and the definition of DUS is formulated by referring to UPOV1991 text. The Chinese of 23 nd 1999 adds the 1978 text of the UPOV convention to become the 39 th member of the UPOV, and the new plant variety right application starts to be received in the same year. The original agriculture department establishes 1 DUS test total center and 14 branch centers nationally on the 9 th and 29 th 2000 th, and specifically bears the DUS test work of new plant varieties. The new revised seed law first brings plant new variety protection into the variety management range at 11.2015, 4, and the DUS test is defined as a precondition for variety approval and registration. With the increasing workload of DUS testing, the original agriculture department subsequently builds 13 sub-centers and 3 testing stations. Up to now, a total of 159 plant protection catalogues are issued by the agricultural rural department, 5 crops such as corn are listed in the variety approval catalogues, and 29 crops such as potato are listed in the variety registration catalogues.

Although UPOV issues a complete set of DUS test guidelines, no specific method of operation is provided. The simple and convenient software for statistical analysis of the DUST in the United kingdom and the simple software for GAIA variety management in France are the only two types of software which are shared freely in the UPOV range at present, but the simple software only relates to the mathematical statistical analysis function and has complex operation, and the simple software for statistical analysis of the variety description function and has strong subjectivity in weight setting, and both the simple software and the simple software have limitations in use.

In recent years, data analysis methods of the country representatives are exchanged in annual meeting of UPOV technical work groups, but the country representatives are mainly concentrated in the aspect of statistical analysis, have scattered functions and many operation steps, particularly lack of quality control methods of original data, so that analysis results are unstable and analysis efficiency is low.

In the past, the inventor continuously and intensively researches, invents a set of comprehensive DUS analysis method, and can perform summarized analysis on data of different places and years by only needing a set of fixed character parameters to obtain scientific and accurate test results, and meanwhile, the analysis speed is greatly increased.

Disclosure of Invention

The invention develops a set of plant variety DUS test and data evaluation and analysis method for efficiently analyzing plant variety DUS test data, improving test level and evaluation quality of DUS test and providing comprehensive and accurate basis for DUS judgment. Furthermore, the data analysis method developed by the present inventors can be conveniently implemented by means of the EXCEL program.

In view of this, there is provided a plant variety DUS testing method comprising performing a plant growing test, collecting and processing test data in a unified format, and performing DUS analysis, the variety types in the plant growing test including a test variety, a standard variety, and optionally an approximate variety; the method is characterized in that the test data are collected and processed in a unified format, and the method comprises the following steps: preparing a unified form, setting a unified data format (such as data type), 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 objective fact is present, the abnormal data is retained and the following steps are continued.

The plant growing test may be performed only for one period of test. Preferably, the plant growing test is performed in two or more phases. One complete cycle of propagation for an annual or biennial plant is the time from sowing to harvesting, and for a perennial plant the time from germination to harvesting for the year of normal flowering and fruiting. Plant growth is greatly affected by temperature, rainfall and illumination, and the plant growth can show great differences among years, so that variety descriptions and differences can not be accurately judged in one year. Therefore, two growth cycles are generally required for the DUS test, usually 3 growth cycles are required for crops or varieties (such as pasture) with poor consistency and small variety-to-variety variation, and 1 growth cycle can also end the DUS test for crops or varieties (such as butterfly orchid) with obvious variety variation in asexual propagation and controlled environment planting (greenhouse).

The stability test is generally carried out by taking seeds of the same variety and different generations for planting test analysis. If the character expression states of the next generation seed and the previous generation seed are consistent and all have consistency, the variety is indicated to have stability. If a variety has consistency in a test, this means stability.

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

The known variety refers to a plant variety which has received an application, or has passed variety approval, variety registration, new variety protection, or has been sold and promoted. The varieties to be tested are varieties which are applied for protection, approval or registration of the right of the varieties, or varieties to be evaluated are selected from the market. Approximate variety: the method is characterized in that the variety which is screened from a variety library for specificity test and is similar to the variety to be tested in phenotype or molecular characteristics and needs to be further verified in field planting tests. Standard variety: refers to known varieties used in planting trials to evaluate environmental impact and to indicate the status of trait expression.

Traits can be classified into quality traits (QL), pseudo quality traits (PQ) and quantitative traits (QN) by expression type. Visual inspection (V) and measurement (M) are classified by observation type. The group (G) and the individual (S) are classified by record type. Combinations by observation type and record type can be classified into group Visual (VG), group Measurement (MG), individual Visual (VS), individual Measurement (MS).

Expression status: in plant variety DUS test guidelines or standards, the expression range of each test trait is divided into a series of expression states. To facilitate defining the trait and specification description, each expression state is assigned a corresponding numerical code to facilitate test data recording, processing, and breed description.

Preferably, the step of making a unified table includes making a unified parameter table, and the parameter table includes at least the following fields of parameters: code, standard value, expression state, standard variety, character number, character name and numerical type; preferably, the parameter table further 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, and photo.

The character number, the code, the expression state, the standard variety, the character name, the expression type, the observation time, the quantity unit, the numerical value type, the maximum value and the minimum value parameters can be preset according to the DUS test guideline.

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

The actual measurement value of the standard variety is the data measured in the test.

The code index is to set an identification code for each code, and can be set to be formed by combining 'character number 10000+codes', so that the corresponding information can be conveniently extracted through the character number and the code later.

The hierarchical value index is another identification code set for each code, and can be set to be formed by combining 'character number 10000+hierarchical value', so that the original value can be conveniently converted into the code by the interval method.

The classification value is used for setting a classification section of the code corresponding to the original value, and is the minimum value of the classification section corresponding to each code.

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.

Grouping: used in the grouping procedure. Grouping basis is grouping trait, described in DUS test guidelines. However, considering the actual use effect, quality traits and/or false quality traits which have discrete expression states, are easy to distinguish and can be accurately observed can be reselected as grouping traits.

The weights are used in calculating variety value and are empirically set. For example, the growth period is important, the weight is 3, the character is less important, the weight is 1, the method is applicable to giving scores in the evaluation of the variety value, and the significance of variety planting is evaluated from the technical point of view.

The threshold value is used for screening approximate varieties by a threshold value method. Is empirically set according to the variation of the data for many years. For example, the quality trait is set to 0 in advance, the false quality trait is set to 1, and the quantitative trait is set to 2. If the preset threshold value is found to be unsuitable after the sample is added or the technology is improved, the proper adjustment can be performed.

Photograph: for each photographed object type, a specific character (for example, a numerical number) is manually assigned in advance, and the photographs are named by the character, for example, the corn photographs are divided into five types of photographs of seedlings, plants, tassels, flowers and ears, the photographs are numbered by 1, 2, 3, 4 and 5, and the photographs are named by the numbers, for example, 1.Jpg, which indicates the photographs of seedlings.

For facilitating analysis and processing, storage management of photos can be optimized, folders of all levels are established according to a unified method, for example, photos\corn\2019\variety names are stored in the folders of the corresponding variety names.

According to the corresponding relation, the characters (such as numbers) for naming the photos are input into the photo fields of the corresponding characters and are linked with the corresponding photos.

In order to interface with other systems, there are cases where it is necessary to uniformly modify the photo name or folder name. At this time, a table is made according to the field formats of the old name, the file type, the file address and the new name, the new name of the photo is input, and the file or the file folder can be renamed in batches by linking the file folder and/or the photo through a program.

As an example, as shown in table 1.

TABLE 1

Regarding the setting of the data format (e.g., data type), the numerical range, and optionally the numerical units, etc., the data type can be set using a table or database self-contained function, with EXCEL cells being an example, the data type being any number, integer, decimal, sequence, date, time, text length, etc. For code-based visual inspection data, sequences can be selected and allowed code values can be filled in the data sources; for continuous measurement data, the decimal may be selected and the data sources filled with allowable minima and maxima; for discrete measurement data, integers can be selected and the data sources filled with allowable minimum and maximum values; for date type data, a date can be selected, and a start date and an end date can be filled in a data source; for the color chart type data, the text length can be selected, and the data source can be filled with the minimum length 4 and the maximum length 5, etc. The numerical range is exemplified by the corn plant height, which can be set to 30-500, and the numerical units, such as cm, are set according to the needs.

Preferably, the data is collected by adopting a table in a uniform format; for example, the data is collected using a horizontal data table or a vertical data table, preferably, the data is collected using a horizontal data table; 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 duplication can occur.

The format of the vertical data table is as follows: and horizontally arranging the fields of the sample numbers of the individual plants to be tested, varieties, tests, characters and the same characters, and taking the character numbers as data vertical rows.

In each table, the variety marked by "yes" under the field to be tested represents the variety which needs to be tested and needs to give an analysis report, and other varieties are represented by "no", such as standard varieties and similar varieties, are not tested and evaluated varieties, and do not need to give an analysis report.

Preferably, in the plant variety DUS test method described above, the testing for validity of test data further includes testing for MS data of a plurality of samples collected in the test using a BoxPlot method (BoxPlot method) and/or a 3σ method (triple standard deviation method); if abnormal value appears, then automatic identification is carried out. If the value is still abnormal by two methods, the original record or the field sample needs to be checked manually, for example, the original record or the field sample belongs to input errors and is directly corrected. In the case of objective facts, if there are few (e.g., two or less) abnormal values that cannot explain the cause, the abnormal values can be corrected by providing an average of the values before and after the program (other calculation methods such as the whole average value, maximum likelihood method to estimate the missing values are not suitable for correcting the DUS test data, and the test results of the plants in the vicinity of the plant are used to estimate the missing values, so that the actual test situation can be reflected more). If the abnormal value is large, the consistency may be checked by a relative variance method or a cou method without processing. Outliers may also be caused by environmental or sampling patterns, such as uneven ground, inconsistent environments, or sampling without excluding marginal plants. This requires optimization of the test design and sampling pattern, and if necessary, expansion of the number of samples.

For the property of the repeated sampling measurement, abnormal values can be more accurately detected by using a BoxPLot or 3 sigma method, wherein the extreme values do not participate in calculation, and the extreme values participate in calculation, and the two methods are complementary.

Preferably, the Boxplot method and/or the 3 sigma method is used for checking by adopting a vertical data table; when the data is collected by adopting the horizontal data table, the data can be converted into the vertical data table through a designed program. The two inspection methods can be performed under the same data format, and various abnormal values are identified by different colors.

Preferably, the collecting and processing of test data in a unified format further comprises the step of converting the trait raw data into code; 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 test or character original data of any multiple tests to perform LSD analysis (the test data of any test or tests can be obtained, the test with the largest variety number is calculated once, and calculation is not needed for each test) 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;

(2) 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 adjusting the classification with reference to whether the percentages of the classification intervals are uniform 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; preferably, the resulting standard value is accurate to the appropriate decimal point number (e.g., 0 decimal, i.e., rounded);

And selecting one plant variety with the measured characteristic actual measurement value at or near each level of standard values 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. The selection of the standard variety should be selected with consideration to the adaptability of the variety, the availability of propagation material and the representativeness of the expression state.

In different test trials, the standard values are usually kept unchanged, and can be verified in a subsequent step, and when an unsuitable situation occurs, such as a new expression state occurs, readjusted and determined. Standard varieties are also typically kept unchanged and are repeatedly planted in each trial. The actual measurement value of the standard variety and the classification value determined by the same vary with the test.

Based on the grading interval and the codes, the original data of the measurement characters MG and MS in the test are coded to obtain interval codes, and the interval codes can be directly used for plant variety DUS analysis or used for plant variety DUS analysis after further optimizing the interval codes.

Preferably, the frequency distribution analysis is performed in a transverse cross test data table, and when the frequency distribution analysis is performed on the primary test MS character data, the to-be-detected, variety, character number and original values thereof are directly extracted from the transverse cross test data table; when frequency distribution analysis is carried out on the MS character data of two or more tests, the numbers of the to-be-tested varieties and characters are extracted from a transverse data table, the test average value of each character is calculated, and the test average value is transferred into a transverse cross test data table, wherein the format of the transverse cross test data table is as follows: the average value of different tests of the to-be-detected, variety and same character or the original values of different plants are continuously arranged in horizontal rows.

Preferably, in the plant variety DUS test method, in the case where the set of standard values adopted for the traits of MG and MS are not obtained from the current test data, the method further includes a step of checking whether the standard variety behaves in the current test in agreement with the test in which the set of standard values is obtained, the step including: comparing the actual measurement value of the standard variety in the test with the corresponding standard value, dividing the absolute value of the difference value of the actual measurement value and the standard value by the standard value, and if the absolute value is more than 10%, confirming the value as an abnormal value and marking the abnormal value (for example, marking the abnormal value as a specific color); for the abnormal value occurrence condition, manually judging whether the abnormal value belongs to acceptable abnormal conditions, and if a plurality of standard varieties of a certain character have similar changes due to a certain factor, judging the abnormal value to belong to acceptable abnormal conditions; if the variation of the standard variety with abnormal characteristics is inconsistent with that of other standard varieties, the actual measurement value of the standard variety needs to be removed.

Preferably, in the step of checking whether the performance of the standard variety is consistent with that of the test for obtaining the set of standard values, calculating the actual measurement value, standard deviation and sample number of the character in the test in a vertical processing data table, and extracting the actual measurement value of the standard variety into a parameter table for checking; the vertical processing data table can be converted from a vertical data table, and the format of the vertical processing data table is as follows: and (3) processing a plurality of sample values of the same MS character in the vertical data table into an average value, a standard deviation and a sample number, reserving fields of interval codes, known codes, regression codes, optimization codes and expression states, and arranging the fields horizontally.

Preferably, for the characteristics of MG and MS, in the case that the set of standard values adopted is not obtained from the current test data, the method further comprises the step of correcting the classification range by using the standard variety, the step comprising: the first one of the classification values is zero, and the second one is respectively: the sum of (measured value-standard value)/standard value obtained for each standard variety is divided by the number of standard varieties, and 1/2 of the sum of the standard value corresponding to the code and the standard value corresponding to the previous code is added.

Preferably, in the plant variety DUS test method, for quantitative traits, the method further comprises the steps of establishing a linear regression function by using a known code of a known variety (including a standard variety and/or an approximate variety) and a corresponding average value of the known variety in the experiment, substituting an original data average value of the variety to be tested in the experiment into the regression function, and solving a regression code of the original data average value; and analyzing the interval codes, the known codes and the regression codes, and selecting the mode codes, the intermediate number codes or the average number codes of the three as further optimized code data.

Preferably, in the above plant variety DUS test method, when at least two of the interval code, the known code, and the regression code are present, further comprising performing a test step using the code range, the test step comprising: calculating the range from the maximum value to the minimum value in the codes, and carrying out different identifications on the range with different sizes; for example, 1 code for yellow, 2 codes for orange, 3 codes for red, 4 codes and above for purple, manually checking the original data for identified code data, or calling a photo confirmation, and manually modifying the optimized code data as needed.

Preferably, in the plant variety DUS test method, the method further includes comparing a plurality of optimized codes of the test together, checking the codes by using the code range obtained by the maximum value-minimum value, performing different identifications according to the range, for example, displaying the identified code data in different specific colors, manually checking the original data, or retrieving photo confirmation, and manually modifying the codes as required to obtain the comprehensive codes crossing the test;

preferably, the checking of the code using the code range is performed in a vertical cross test data table, and the format of the vertical cross test data table is as follows: 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; calculating the code range of different tests, and carrying out different marks according to the range, for example, displaying in different specific colors, such as yellow (range 1), orange (range 2), red (range 3) and purple (range 4 or more); and for the developed code data, namely codes with differences among tests, manually checking the original data or retrieving photo confirmation, manually modifying the codes according to the requirement, and taking the confirmed codes as cross-test comprehensive codes.

Preferably, in the plant variety DUS test method described above, the method further comprises the step of transferring the transformed code into a variety library; preferably, if the variety or corresponding trait already exists in the variety pool, the result is overlaid; if not, adding varieties in the last row of the variety library or adding characters in the last column, and updating the variety library;

preferably, after two or more tests are completed, photographs of the two or more tests are compared one by one, whether the two or more tests have differences is determined, if so, the reason is checked, and if not, a set of standard photographs is selected and stored in a preset folder, for example, a dus\corn\standard photograph\variety name folder.

Preferably, the method further comprises the step of confirming or correcting the character code by using the photo; preferably, after arranging a certain character code in a sequence from small to large, extracting photos corresponding to the character code for sequential visual comparison; preferably, the operation is performed in a horizontal data table or variety library, the mouse is marked on a certain code character column, 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 characters (preferably digital numbers) named by the photos, the photos are placed in the corresponding position of the next column, the photos are checked in sequence, whether the error of the codes is caused or not is confirmed manually, and finally, the consistency of the codes with reports and the photos is ensured.

The method optimizes the DUS test design, realizes the joint correction of a plurality of test data, and increases the objectivity of the DUS test analysis result. Meanwhile, the method can efficiently realize test data and statistical hypothesis validity check and DUS data analysis by means of the EXCEL program, so that the DUS data analysis work which can be completed in 2 months originally is reduced to be completed in 1 day.

Drawings

FIG. 1 shows a vertical processing data table showing linked photos;

FIG. 2 shows a vertical cross test data table showing linked photographs;

FIG. 3 shows a diagram of two experimental comparative interface examples;

FIG. 4 is a diagram showing an exemplary interface for photo validation in a variety library;

FIG. 5 is a diagram showing an example of a comparison interface between a test variety and an approximate variety in a variety library;

Detailed Description

The following describes specific embodiments of the present invention in detail. It should be understood that the detailed description and specific examples, while indicating the invention, are intended for purposes of illustration only and are not intended to limit the invention.

The test crops are corn, 149 varieties, 147 varieties and 2 standard varieties. Each planting was performed once in the first and second years. Each cell is 5 m long and 2.4 m wide, four rows of plants are planted, the plant row spacing is 30cm multiplied by 60cm, two-seed sowing is carried out, two leaves and one core are used for thinning, at least 80 plants are reserved in each cell, and two repetition is arranged. The field management measures are the same as the field production. When data are collected, VG and MG characters only collect one value, VS and MS characters collect 20 single plant values, and five kinds of seedlings, plants, tassel, filaments and ears are photographed by photographs, and the samples are all from the same cell. Five photographs of seedlings, plants, tassel, filament and ear were taken throughout the growth period.

1. Setting a parameter table of corn according to DUS test guidelines

The following only take three characters as examples because of the large number of characters, and the parameter table settings are shown in table 2:

TABLE 2

In table 2, parameters such as a property number, a code, an expression state, a standard variety, a property name, an expression type, an observation time, a quantity unit, a numerical value type, a maximum value, a minimum value and the like can be preset according to a DUS test guideline, and a 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.

2. The horizontal data table records the original data

The recording format is shown in table 3 below. Only the first 16 traits and the first 14 varieties are listed because of the large variety and number.

TABLE 3 Table 3

Performing horizontal data acquisition according to a uniform horizontal data format, and manufacturing a horizontal data table: and transversely arranging according to the fields of the to-be-detected, variety, test and character numbers. When a plurality of individual sample values are measured for the same trait, the same trait number is continuously repeated in horizontal rows. For example, in table 3, 20 individual samples were measured for the trait numbered 16, with trait numbered 16 being in consecutive horizontal rows.

In table 3, the variety identified by "yes" in the field to be tested indicates the variety that needs to be tested and needs to give an analysis report, and other varieties are indicated by "no", such as standard varieties and similar varieties, and are not the variety to be tested and evaluated, and need not give an analysis report for them.

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

3. Data verification

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

And checking the horizontal data based on the data formats such as the data types, the numerical ranges such as the minimum value and the maximum value, the numerical units such as the days and the cm in the part 1 parameter table.

In the test result, an abnormal value is identified with a specific color, for example, red. For example, based on the setting of the parameter table of the 1 st part, the numerical range of the property 2 should fall within the range of 1-5, and 6 recorded during data acquisition is an abnormal value; the numerical range of the character 16 is 10-70, and 4 recorded during data acquisition is an abnormal value. These outliers are automatically displayed as red by the design program.

(2) Test data validity test using Boxplot and 3 sigma methods

And converting the horizontal data format into the vertical data format, and manufacturing a vertical data table. Test data validity tests were performed using the Boxplot method and the 3σ method.

The transformed vertical data table is shown in table 4. And horizontally arranging the fields of the sample numbers of the individual plants to be tested, varieties, tests, characters and the same characters, and taking the character numbers as data vertical rows.

TABLE 4 Table 4

The same vertical data table can be used for two calculations and various outliers can be identified with different colors.

The calculation results of the Boxplot method are shown in Table 5 (1.5 times pitch in yellow, 3 times pitch in red).

TABLE 5

The 3 sigma method calculation results are shown in table 6 (yellow is 2 times standard deviation, red is 3 times standard deviation).

TABLE 6

Through two methods, the test is still an abnormal value, and the original record or the field sample needs to be checked manually, for example, the original record or the field sample belongs to input errors and is directly corrected. In the case of objective facts, if there are few (e.g., only two or less) outliers for which the cause cannot be explained, the procedure provides a front-to-back value average. If the abnormal value is more, the consistency is checked by a relative variance method or a COYU method without processing.

4. Frequency distribution analysis

(1) When the frequency distribution analysis is carried out on the MS character data of the primary test, the to-be-tested, variety, character numbers and original values thereof can be directly extracted from the transverse data table to the transverse cross test data table; (2) When frequency distribution analysis is carried out on the MS character data of the two tests, the numbers of the to-be-tested, the variety and the character can be extracted from a transverse data table, the test average value of each character is calculated, and the test average value is transferred into a transverse cross test data table, wherein the format of the transverse cross test data table is as follows: the average value of different tests of the to-be-detected, variety and same character or the original values of different plants are continuously arranged in horizontal rows.

The cross-row test data table is shown in table 7.

TABLE 7

Frequency distribution analysis and determination of standard value and gradation value

The numerical trait frequency distribution calculation results are shown in table 8.

TABLE 8

Traits (3)	16	17	18	19	22	25.2	26.2	27.2	29.2	30.2	31.2
												Total mean value of	40.68	29.72	7.956	25.19	10.58	103	276.6	0.372	19.96	4.599	16.26
Sum total	57769	42198	11297	35771	15019	1E+05	4E+05	528	28338	6530	23088
												Total variance of	18.24	15.54	9.829	13.03	0.92	276.7	711	0.002	3.612	0.068	3.169
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	Intermediate value	Minimum value	Quantity of	Frequency of	Minimum value of	Average value of	Maximum value
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	Intermediate value	Minimum value	Quantity of	Frequency of	Minimum value of	Average value of	Maximum value
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 raw data of all varieties is taken as a central standard value, and 2 times of LSD is taken _0.05 Setting standard values of all levels for the level differences, and setting a section formed by extending 1/2 level differences to two sides by taking the standard values of all levels as the center, wherein the section is set as a grading section of each code; the minimum value of each interval is used as a grading value; and determining the grading value and the grading interval of each grade, and counting the number and the percentage of varieties in each grading interval. And judging whether the total interval coverage is less than level 3 or greater than level 9 according to the statistical result. Traits less than grade 3 are not suitable for DUS testing to be rejected; traits greater than grade 9, modulating LSD _0.05 And adjusting the classification by referring to whether the percentages of the intervals of each stage are uniform, so that the classification range is within 9 stages; the characters between 3 and 9 grades can be applicable when 1-2 grades are left at the two ends respectively so that new varieties appear in the future; thereby a more rational grading is obtained. 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, re-determining the standard value, and rounding the standard values of all levels. And judging whether the code position of the standard variety is proper or not according to the performance of the standard variety in the test, if not, moving the standard variety to the proper code position, and carrying out integral translation on the standard value according to the actual measurement value of the standard variety so as to enable the actual measurement value of the standard variety to be close to the corresponding standard value. Thus, the standard variety is determined and a set of relatively fixed standard values is formed. As shown in table 9.

TABLE 9

5. The vertical data format is converted into a vertical processing data format, a vertical processing data table is manufactured, and code data inspection and optimization are carried out

And converting the vertical data format into a vertical processing data format, namely a to-be-detected, variety, test and character horizontal row, processing a plurality of sample values of the same MS character in a vertical data table into an average value, a standard deviation and a sample number, reserving interval codes, known codes, regression codes, optimization codes and expression state fields, and horizontally arranging the fields. The processed table format is shown in table 10.

Table 10

In the case where the set of standard values used for the characteristics of MG and MS is not obtained from the current test data (for example, the second year test), it is checked whether the standard variety performs the current test in agreement with the test in which the set of standard values was obtained.

In the vertical processing data table, the average value (i.e., actual measurement value), standard deviation, and number of samples of the property in the present test are calculated. And extracting the actual measurement value of the standard variety into a parameter table. If the absolute value of the difference between the measured value of the standard variety and the standard value is divided by the standard value, and the difference is determined to be excessive if the absolute value is greater than 10%, the indication is displayed in a specific color (for example, red). For the case of abnormal values, it is manually determined whether or not the abnormal values are acceptable (quality characteristics in principle do not allow abnormality to occur, and if so, the cause needs to be manually checked and corrected). If a plurality of standard varieties of a certain character have similar changes due to a certain factor, the method belongs to acceptable abnormality and data retention. If the variation of the standard variety with abnormal characteristics is inconsistent with that of other standard varieties, the actual measurement value of the standard variety needs to be removed.

For the characteristics of MG and MS, when the set of standard values adopted is not obtained from the test data, the standard variety is used for correcting the grading range. The first one of the classification values is zero, and the second one is respectively: the sum of (measured value-standard value)/standard value of each standard variety is divided by the number of standard products, and 1/2 of the sum of the standard value corresponding to the code and the standard value corresponding to the previous code is added.

And calculating a grading value, a grading value index and a grading interval corresponding to each code according to the standard value and the actual measurement value, and calculating the code of each property of each variety to obtain an interval code.

For quantitative traits, if a known variety with data in the variety library exists in the test, the corresponding code of the known variety in the variety library is referred to the known code in the vertical processing data table. And establishing a linear regression relation between the average value of the known varieties in the test and the corresponding known codes, substituting the average value of the original data of other varieties in the test into the linear regression relation, and calculating the regression codes of all varieties. And analyzing the interval codes, the known codes and the regression codes, and selecting the mode codes, the intermediate number codes or the average number codes of the mode codes, the intermediate number codes and the average number codes as further optimized code data.

When at least two codes of an interval code, a known code and a regression code exist, the code range is adopted for checking. The extreme difference is calculated from the maximum value minus the minimum value among these codes, and a specific color is displayed in the extreme difference size, for example, 1 code for yellow, 2 codes for orange, 3 codes for red, 4 codes for purple, and the like. For the developed code data, the original data is manually checked, or photo confirmation is invoked, and the code can be manually modified as needed. By designing the program, the photo can be quickly taken for confirmation. And placing a mouse on the code of which the photo needs to be adjusted, linking the photo through a program, and displaying the photo at the blank position of the same row of the code. As shown in fig. 1.

6. Converting the vertical-row processing data format into a vertical-row cross-test data format, and determining a cross-test comprehensive code

And converting the vertical-row processing data format of the two-year test into a vertical-row cross test data format, and manufacturing a vertical-row cross test data table as shown in the following table. The vertical row cross-year data table format is: the average value, standard deviation, sample number, code and expression state in each annual test are displayed in parallel, and the average value, standard deviation, sample number and optimized code average value in two annual tests are calculated and displayed in parallel. Wherein the average value of the optimization code is rounded (directly after the decimal point is removed, the value is rounded off or rounded off). The individual year code range is calculated, and the individual year codes are displayed in different specific colors, for example, yellow (range 1), orange (range 2), red (range 3), and purple (range 4 or more) according to the range size. For the developed code data (code with differences between years), the original data is manually checked, or photo confirmation is invoked, and the code can be manually modified as needed. The validated code acts as a cross-trial composite code. By designing the program, the quick photo taking and confirmation can be realized. And placing a mouse on the code of which the photo needs to be adjusted, linking the photo through a program, and displaying the photo at the blank position of the same row of the code. As shown in fig. 2.

7. Transferring cross-test comprehensive codes into variety libraries

The integrated code data from the cross-test data table was transferred to the variety library as shown in table 11. For the existing varieties or characters in the variety library, code data is directly covered; the variety or character not in the variety library is added with a field in the last row (variety) or the last column (character) of the data, and the comprehensive code data is imported. Thus, the data of the variety library is updated and accumulated year by year.

TABLE 11

To be measured	Variety of species	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26
																												Is that	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
Is that	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
																												Is that	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
Is that	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
																												Is that	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
Is that	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
																												Is that	North farmer 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
Is that	North farmer 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
																												Is that	North farm 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
Is that	North agriculture 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
																												Is that	Jade 88	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
Is that	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
																												Whether or not	Cartesian 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
Is that	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 finishing

After two years of testing, two years of photographs were compared. It is determined whether there is a difference in the two-year photographs. If the difference exists, checking the reason, if the difference exists, selecting a set of standard photos, and storing the standard photos into a DUS/corn/standard photo/variety name folder. An example of a comparison interface is shown in fig. 3.

9. Batch retrieving photo validation codes after code ordering

In a horizontal data table or a variety library, a mouse is marked on a certain code character column, a program is automatically sequenced according to the size of the code, a photo type number corresponding to the character is preset under a character photo field in a parameter table of the 1 st part, photos of each variety corresponding to the character are extracted in batches, the photos are placed at the corresponding position of the next column, the photos are checked in sequence, and whether the code is wrong or not is confirmed manually. Finally, the codes with reports are drawn up to be consistent with the photos. An exemplary diagram of an interface for photo validation in a variety library is shown in FIG. 4.

10. Specificity analysis

(1) Similar variety screening in variety library

And in the finally formed variety library, firstly, sorting all varieties in groups according to grouping information set in the parameter table. The varieties can be quickly and preliminarily divided into a plurality of groups. If there is only one variety within a group, then that variety need not be compared to other varieties.

Then, the varieties are analyzed for the approximation degree by using a differential character number accumulation method, a character number accumulation method with the difference larger than a threshold value, a correlation coefficient method or a minimum distance method, and general reminding and special warning numerical intervals are respectively set as shown in table 12.

Table 12

Taking correlation coefficient as an example, in a result area, the horizontal row is the variety to be detected, the vertical row is all varieties, and the correlation coefficient is displayed in yellow more than 90% and in red more than 95%.

(2) Further confirmation by variety photo comparison

Deleting all data in the fields to be detected in the variety library table, filling "Yes" before the variety to be detected which needs to be further compared, and filling "No" before the approximate variety analyzed in the step (1). According to the input information, pictures of the variety to be tested and the similar variety are sequentially called through a program, displayed side by side, and manually and rapidly checked and confirmed whether the difference exists or not. An example of a comparison interface between the test variety and the similar variety photograph is shown in FIG. 5.

(3) Calling original data

If the similar variety has no difference with the variety to be tested, further calling the average value and code (MS character also including standard deviation) of all the measured characters of the two varieties from the vertical processing data table or the vertical cross test data table, putting the average value and code into a data comparison table, and comparing the average value and code side by side, wherein the data with larger difference are checked as shown in a table 13.

TABLE 13

(4) And aiming at the characteristics of MG and MS, continuously taking single test original data or double test summarized data of two varieties to carry out T test or COYD analysis. As shown in tables 14 and 15.

TABLE 14

TABLE 15

(5) For the VS trait, in the vertical row and column data sheet, the specificity of the varieties to be tested and the similar varieties is analyzed by the Pearson chi-square test method.

The original data can be statistically converted into a vertical row column data format from the vertical row data format, and the format of the vertical row column data table is as follows: the test, breed, trial, trait, code fields are 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. The field data can also be collected directly according to the vertical row column data format.

Table 16

To be measured	Variety of species	Test	Traits (3)	1	2	3
							Is that	C01	2018	1	34	6	6
Whether or not	R01	2018	1	12	23	9
							Whether or not	R02	2018	1	6	20	19
Whether or not	R03	2018	1	1	18	9
							Whether or not	R04	2018	1	7	22	15
Is that	C02	2018	1	9	3	34
							Whether or not	R05	2018	1	4	8	34
Whether or not	R06	2018	1	1	11	34

The pearson chi-square test results are shown in table 17.

TABLE 17

To be measured	Variety of species	Test	Traits (3)	1	2	3		C01	Results	C02	Results
												Is that	C01	2018	1	34	6	6	1			2E-08	D
Whether or not	R01	2018	1	12	23	9		3E-05	D	3E-07	D
												Whether or not	R02	2018	1	6	20	19		4E-08	D	0.0002	D
Whether or not	R03	2018	1	1	18	9		2E-08	D	5E-07	D
												Whether or not	R04	2018	1	7	22	15		2E-07	D	2E-05	D
Is that	C02	2018	1	9	3	34	2	2E-08	D
												Whether or not	R05	2018	1	4	8	34		3E-10	D	0.1227	ND
Whether or not	R06	2018	1	1	11	34		5E-12	D	0.0041	D

(6) When there are only two expression states in (5), the specificity can be analyzed by the Fisher exact test method with higher accuracy.

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

TABLE 18

To be measured	Variety of species	Test	Traits (3)	1	2		C01	Results	C02	Results
											Is that	C01	2018	1	34	6				0.2295	ND
Whether or not	R01	2018	1	12	23		6E-06	D	0.0146	ND
											Whether or not	R02	2018	1	6	20		5E-07	D	0.0033	D
Whether or not	R03	2018	1	1	18		3E-09	D	9E-05	D
											Whether or not	R04	2018	1	7	22		4E-07	D	0.0033	D
Is that	C02	2018	1	9	3		0.2295	ND
											Whether or not	R05	2018	1	4	8		0.0011	D	0.0436	ND
Whether or not	R06	2018	1	1	11		2E-06	D	0.0013	D

11. Discrete data consistency analysis

The consistency analysis of discrete data adopts a UPOV abnormal strain method, the method reforms a display interface, and is convenient for data recording and batch calculation, and the interface is shown in a table 19.

TABLE 19

To be measured	Variety of species	Test	Traits (3)	Total plant number	Abnormal plant number		Abnormal plant method	Consistency of
									Whether or not	R01	2017	1	20	3		0.978991644	NU
Whether or not	R02	2017	1	30	3		0.996682291	NU
									Whether or not	R03	2017	1	40	3		0.992502637	NU
Whether or not	R04	2017	1	50	3		0.810798075	U
									Whether or not	R05	2017	1	60	3		0.73146611	U
Whether or not	R06	2017	1	70	3		0.649236912	U
									Whether or not	R07	2017	1	80	6		0.966693618	NU
Whether or not	R08	2017	1	90	6		0.946079612	U
									Whether or not	R09	2017	1	100	6		0.919162871	U
Whether or not	R10	2017	1	110	6		0.886011935	U
									Whether or not	R11	2017	1	120	6		0.847081858	U
Is that	C01	2017	1	130	6		0.803137839	U
									Is that	C02	2017	1	140	9		0.974122695	NU
Is that	C03	2017	1	150	9		0.962190432	NU
									Is that	C04	2017	1	160	9		0.946973552	U
Is that	C05	2017	1	170	9		0.928236031	U
									Is that	C06	2017	1	180	9		0.905863816	U
Is that	C07	2017	1	190	9		0.879872217	U

The P value and the result are calculated according to three values of the population size, the abnormal plant number and the population standard.

12. Continuous data consistency analysis-relative variance method;

data for one year can be analyzed for consistency using the relative variance method, with data acquisition and analysis interfaces as shown in table 20.

Table 20

To be measured	Variety of species	Test	Traits (3)	Total plant number	Standard deviation of		Relative variance	Allowing relative variance	Consistency of
										Whether or not	R01	2017	1	20	1.5		1.012269939	1.878311735	U
Whether or not	R02	2017	1	20	2.2		1.484662577	1.878311735	U
										Whether or not	R03	2017	1	20	2.3		1.552147239	1.878311735	U
Whether or not	R04	2017	1	20	1.4		0.944785276	1.878311735	U
										Whether or not	R05	2017	1	20	0.5		0.337423313	1.878311735	U
Whether or not	R06	2017	1	20	1.4		0.944785276	1.878311735	U
										Whether or not	R07	2017	1	20	1.4		0.944785276	1.878311735	U
Whether or not	R08	2017	1	20	1.4		0.944785276	1.878311735	U
										Whether or not	R09	2017	1	20	1.4		0.944785276	1.878311735	U
Whether or not	R10	2017	1	20	1.4		0.944785276	1.878311735	U
										Whether or not	R11	2017	1	20	1.4		0.944785276	1.878311735	U
Is that	C01	2017	1	20	1.7		1.147239264	1.878311735	U
										Is that	C02	2017	1	20	4.1		2.766871166	1.878311735	NU
Is that	C03	2017	1	20	1.53		1.032515337	1.878311735	U
										Is that	C04	2017	1	20	1.47		0.99202454	1.878311735	U
Is that	C05	2017	1	20	1.41		0.951533742	1.878311735	U
										Is that	C06	2017	1	20	3		2.024539877	1.878311735	NU
Is that	C07	2017	1	20	1.29		0.870552147	1.878311735	U

13. Continuous data consistency analysis-COYU method

The two-year data are obtained from a transverse cross test data table by adopting a COYU method, and an analysis interface is shown in a 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.