Method for selective breeding of fine-wool sheep genome

Abstract

The invention belongs to the technical field of breeding, and particularly relates to a fine-wool sheep genome selective breeding method. According to the method, genetic typing is carried out on each sample by using a fine-wool sheep 50K SNP liquid phase chip, statistical analysis is carried out through genetic typing and phenotype data, RRBLUP and SSGBLUP are used for training population genotype data and phenotype data and calculating prediction accuracy, and finally, the net wool amount, wool length, wool fiber diameter and weight are selected as four breeding target characters. A genome breeding value (GEBV) of the weight, the wool length, the wool fiber diameter and the net wool amount of the fine-wool sheep is used as a denominator, the standard economic value of each character is used as a molecule, a fine-wool sheep genome comprehensive selection index estimation model (GPI) is formulated, and selective breeding of the fine-wool sheep genome is performed according to the GPI numerical value. A genome selection strategy with high accuracy and low cost is provided, and the method has important significance on high-quality development of fine-wool sheep breeding industry and industry.

Description

Technical Field

[0001] The invention belongs to the technical field of breeding, and in particular relates to a method for genome selection breeding of fine-wool sheep. Background Art

[0002] Genomic selection is a method of predicting the genetic value of animals based on genotyping data of genetic markers distributed on the genome, thereby selecting ideal individuals. Breeders can increase the speed of genetic improvement through early selection.

[0003] With the continuous improvement of reference genomes, the reduction in the cost of sequencing technology and the continuous upgrading of commercial SNP chips, genomic selection is currently the mainstream method of animal breeding and has been widely used in livestock and poultry such as chickens, beef cattle, pigs, goats and sheep. The fastest growing application is in dairy cows, which have relatively high individual value.

[0004] Fine-wool sheep breeding in my country is still in the stage where phenotypic selection is the main focus. There is not only a lack of large-scale, high-quality fine-wool sheep genomic selection reference populations, but also a lack of specialized chips for fine-wool sheep genomic selection breeding. This leads to higher genotyping costs and lower economic returns, limiting the innovative application of genomic selection in fine-wool sheep breeding.

[0005] Therefore, developing specialized chips for genomic selection and breeding of fine-wool sheep as a low-cost means of genotyping and exploring high-accuracy and low-cost genomic selection strategies are of great significance to the high-quality development of the fine-wool sheep breeding industry. Summary of the invention

[0006] In view of the above technical problems, the present invention provides a method for genomic selection breeding of fine wool sheep, which specifically includes the following contents:

[0007] A method for genomic selection breeding of fine wool sheep, the method comprising the following steps:

[0008] (1) Calculate the genomic estimated breeding values ​​(GEBV) of four target traits of fine wool sheep, namely, weight, wool length, wool fiber diameter, and wool mass;

[0009] (2) Calculate the standard economic value (VσA) of the four target traits of fine wool sheep: weight, wool length, wool fiber diameter and net wool yield;

[0010] (3) Constructing a comprehensive selection index estimation model for fine wool sheep genome: as shown in formula (I),

[0011]

[0012] Where, GPI: genome comprehensive selection index; GEBV: CFW : Genomic Estimated Breeding Value for Gross Weight; GEBV SL: Genomic Estimated Breeding Value for Wool Length; GEBV FD : Genomic Estimated Breeding Value for Wool Fiber Diameter; GEBV WT : Genomic estimated breeding value of body weight; VσA CFW : Net gross genome standard economic value; VσA SL : Economic value of wool length genomic standard; VσA FD : Economic value of wool fiber diameter genome standard; VσA WT : Economic value of the body weight genomic standard;

[0013] (4) Select fine-wool sheep individuals with high GPI values ​​for breeding.

[0014] Preferably, the GEBV calculation model is as shown in formula (II):

[0015]

[0016] In the formula, GEBV i is the genomic estimated breeding value of trait i, Z j is the genotype vector at site j; b j is the estimated effect size of site j.

[0017] Preferably, the VσA calculation model is as shown in formula (III):

[0018] v σAi = v i ×σ A , (III);

[0019] In the formula, v σAi is the standard economic value of trait i; v i is the economic value of trait i; σ A For v i The corresponding genetic standard deviation.

[0020] Preferably, the v i The difference method is used for calculation, and the formula is shown in formula (IV):

[0021] v i =∑ΔR ij -∑ΔC ij , (IV);

[0022] In the formula, ΔR ij is the jth output component of trait i; ΔC ij is the jth input component of trait i; when calculating the economic value of the target traits of Alpine Merino sheep breeding, it is calculated based on the lambing cycle of one ewe.

[0023] Preferably, the GEBV iSelect either the SSGBLUP or RRBLUP method for calculations.

[0024] Preferably, the SSGBLUP method is:

[0025] Based on the kinship matrix A and the genome relationship matrix G, a new matrix H is established by integration:

[0026]

[0027] In formula (V), A and G are A matrix and G matrix respectively, and subscripts 1 and 2 are individuals without genotype and individuals with genotype respectively; since G cannot be inverted when it is a singular matrix, G is defined as:

[0028] G ω = (1-ω)G + ωA 22 , (VI);

[0029] Then the H inverse matrix can be transformed into:

[0030]

[0031] In formula (VII), ω is the weighting factor, i.e., the proportion of polygenic genetic effects;

[0032] After constructing the H matrix, it solves the system of equations for the mixed model:

[0033]

[0034] Preferably, the RRBLUP method is: calculating each marker effect value, then accumulating the effect values, and then obtaining the breeding value;

[0035] Assume that the variance of the effect values ​​of all SNP sites is equal, that is:

[0036] σ g1 2 =σ g2 2 =…=σ gM 2 , (Ⅸ);

[0037] The effect size of the SNP can be obtained by solving the following mixed model equations:

[0038]

[0039] Where M is the number of SNP sites, h 2 The chivalrous heritability of the target trait.

[0040] Preferably, the genetic variance is estimated using the kin.blup function of the RRBLUP package in R software. and error variance The heritability calculation formula is:

[0041]

[0042] Preferably, the accuracy evaluation formula of GEBV is:

[0043]

[0044] In the formula, y represents the corrected phenotypic value.

[0045] Preferably, the y calculation formula is:

[0046]

[0047] Where X is the fixed effect correlation matrix, a represents the fixed effect vector, and e represents the random residual vector.

[0048] The beneficial effect of the present invention is as follows: the present invention performs genotyping on each sample by utilizing a 50K SNP liquid phase chip of fine-wool sheep (refer to patent CN117025786A), statistically analyzes the genotyping and phenotypic data, and uses RRBLUP and SSGBLUP to train the population genotypic data and phenotypic data and calculate the prediction accuracy, and finally selects net wool quantity, wool length, wool fiber diameter and body weight as four breeding target traits, and uses the genomic breeding value (GEBV) of fine-wool sheep weight, wool length, wool fiber diameter and net wool quantity as the denominator and the standard economic value of each trait as the numerator to formulate a fine-wool sheep genome comprehensive selection index estimation model (GPI), and performs selective breeding of the fine-wool sheep genome according to the GPI numerical value; a genomic selection strategy with high accuracy and low cost is provided, which is of great significance to the high-quality development of fine-wool sheep seed industry and industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 Sample core SNP site deletion and mutation statistics;

[0050] Figure 2 Core SNP annotation statistics;

[0051] Figure 3 Heat map of phenotypic correlation coefficients. DETAILED DESCRIPTION

[0052] The present invention is further explained with reference to the detailed description of the specific embodiments below, but the following embodiments are merely illustrative of the technical solutions of the present invention and do not limit the technical solutions of the present invention. Unless otherwise specified, the technical means used in the following embodiments are conventional means well known to those skilled in the art, the bioinformatics software and products used are also commercially available, and the experimental processes and methods are also conventional methods known in the art. The sources of the materials used, the trade names, and the components thereof that need to be listed are all indicated when they first appear, and the same reagents used thereafter are the same as those indicated for the first time unless otherwise specified.

[0053] In addition, it should be noted that the site combination and application provided by the present invention are completed by the inventors of this application through arduous creative labor and optimization work.

[0054] It should be noted that the economic traits of fine wool sheep described above were objectively tested in accordance with relevant national standards at the Animal Fur and Products Quality Supervision and Inspection Center of the Ministry of Agriculture and Rural Affairs (Lanzhou, China). Phenotypic data such as weight and shearing amount were collected from the fine wool sheep production performance measurement site and shearing site using well-known methods.

[0055] Example 1 Method for genomic selection breeding of fine wool sheep

[0056] 1. Construction of a fine-wool sheep reference group

[0057] Building a reference group is the first crucial step in implementing genomic selection breeding. Building a reference group requires a large amount of phenotypic and genotypic data support. The accuracy of breeding value estimation is closely related to the size and structure of the reference group. More than 2000 fine-wool sheep core groups are selected, all of which are adult ewes, and blood and wool samples are collected. All individuals have complete phenotypic records and pedigree archives. This embodiment selects 2178 alpine merino sheep from five core groups of the Zhangye Experimental Station of the National Wool Sheep Industry Technology System (Gansu Sheep Breeding Technology Promotion Station), all of which are adult ewes, and blood and wool samples are collected according to the experimental design. All individuals have complete phenotypic records and pedigree archives. By using the fine-wool sheep 50KSNP liquid phase chip (refer to patent CN117025786A) to perform genotyping on each sample, the genomic breeding value (GEBV) of each individual is calculated, and the construction of a high-quality reference group for alpine merino sheep genome selection is finally realized.

[0058] 2. Genotyping and Data Processing

[0059] Genotyping: A high-throughput DNA extraction kit is used to extract sample DNA. The extracted DNA samples are tested for purity and integrity and the concentration is accurately quantified. DNA that has passed the quantitative quality inspection is randomly physically broken using an ultrasonic crusher, and the peak value of the broken fragment is controlled at 200-300bp. The broken DNA is connected to the A tail after end repair. The DNA fragments with A added are connected to the sequencing adapter using ligase, and then the library is purified and fragments are selected using carboxyl-modified magnetic beads, and the connection products with inserted fragments of 200-300bp are retained. The connection products of the sequencing primers with Barcodes are added to the high-fidelity PCR reaction system for PCR amplification and purification for probe hybridization experiments. Take 500ng of the constructed sequencing library, freeze-dry it, add the probe and hybridization reagent, denature it and incubate it at 65℃ for 2 hours to complete the hybridization reaction. After the hybridization product is washed with washing solution, another round of PCR is performed to complete the construction of the hybridization capture library. After the library construction is completed and quality inspection is carried out, it enters the sequencing stage on the machine.

[0060] Data processing: To ensure high-quality markers to improve the prediction ability of genome selection, after the genotyping data is downloaded, the markers are first filtered for completeness, and the marker genotypes covering more than 80% of the individuals are screened and retained; secondly, heterozygosity filtering is performed to screen and retain markers with a heterozygosity rate greater than 50%; thirdly, minimum allele frequency filtering is performed to screen and retain markers with a minimum allele frequency (MAF) greater than 0.05; finally, multi-allelic variation sites are filtered, that is, multi-allelic variation sites are removed and di-allelic variation sites are retained.

[0061] 2178 alpine Merino sheep samples were captured and sequenced using fine wool sheep 50K SNP liquid phase chip, and finally 30,791,025,906 Raw Reads were obtained, and 30,556,682,363 Clean Reads were obtained after filtering; a total of 4,618,653,885,858 bp of Raw Bases were obtained, and 4,215,263,832,905 bp of Clean Bases were obtained after filtering, with an efficiency of 86.41% to 95.49%, Q20 of 94.7% to 98.81%, and Q30 of 86.64% to 95.83%. A total of 26,052,146,716 Clean Reads were aligned to the reference sequence, and the sample alignment rate was 87.23% to 96.9%. A total of 46178 loci were obtained after genotyping of Alpine Merino sheep using the fine-wool sheep 50K SNP liquid phase chip. 44604 loci were retained after integrity filtering, 43847 loci were retained after heterozygosity filtering, 40549 loci were retained after minimum allele frequency filtering, and 40539 loci were retained after multi-allelic variation site filtering.

[0062] 3. Phenotypic Data Determination and Statistical Analysis

[0063] Six wool traits (fiber diameter, fiber diameter variation coefficient, hair pile length, shearing amount, clean hair rate and clean hair volume) and one body weight trait were selected for genomic selection of fine wool sheep. The three wool traits of wool fiber diameter (FD) (GB / T 10685-2007), fiber diameter variation coefficient (FD_CV) (GB / T 10685-2007) and clean hair rate (CFWR) (GB / T 6978-2007) were tested at the Animal Fur and Products Quality Supervision and Inspection Center of the Ministry of Agriculture and Rural Affairs (Lanzhou, China) according to relevant national standards. The shearing amount data of fine wool sheep were measured on-site in August 2021, and the clean hair volume data was the shearing amount multiplied by the clean hair rate; the body weight (WT) and hair pile length (SL) were determined on-site in June 2021. All phenotypic data were tested for extreme outliers using R software (v3.5.0), and the abnormal samples with extreme phenotypic outliers (i.e., values ​​outside the range of the mean plus or minus 3 times the standard deviation were identified as extreme outliers and removed) were removed according to the 3δ rule for subsequent analysis. Descriptive statistical analysis was performed on the phenotypic data before and after the removal of outliers, including mean, standard deviation, standard error, skewness, and kurtosis. The phenotypic data after the removal of extreme outliers were analyzed by Pearson's method of the cor function in R language (version 4.0.3) for Pearson correlation analysis and plotting.

[0064] The phenotypic data were tested for extreme values ​​and descriptive statistics were performed, as shown in Table 1.

[0065] Table 1 Descriptive statistics of phenotypes after removal of extreme outliers

[0066]

[0067]

[0068] 4. Estimation of genetic parameters of main traits

[0069] The genetic variance was estimated by running the kin.blup function using the RRBLUP package in R software ( genetic variance) and error variance ( The phenotypic and genotypic data of the reference group are used to estimate the population genetic variance and calculate the heritability of each trait. The heritability calculation formula is:

[0070]

[0071] 5. Evaluation of Genomic Prediction Accuracy

[0072] 5.1 Principal Component Analysis

[0073] To eliminate the influence of population genetic structure on the accuracy of genome prediction, PCA analysis was performed based on the filtered SNP markers using GCTA software (v1.92.2).

[0074] 5.2 Model selection

[0075] The five-fold cross-validation method was used to evaluate the estimation accuracy of GEBV. The reference population was randomly divided into five groups of essentially equal size, each of which was used as a validation group in turn, and the remaining four groups were used as training groups to verify the accuracy of GEBV prediction. Finally, the accuracy and standard error were calculated for each trait to reflect the prediction accuracy. The RRBLUP and SSGBLUP models were selected to train the genotype and phenotypic data of the reference population.

[0076] The evaluation formula for GEBV accuracy is:

[0077]

[0078] In the formula, y represents the corrected phenotypic value, h 2 Represents the heritability of a trait.

[0079] The results of using RRBLUP and SSGBLUP to train the population genotype data and phenotype data and calculate the prediction accuracy are shown in Table 2.

[0080] Table 2 Statistics of prediction accuracy of different models for each trait

[0081]

[0082]

[0083] Using the RRBLUP and SSGBLUP models, the phenotype and genotype of the reference population were trained with flock and age as fixed factors, and the genome prediction accuracy was calculated using five-fold cross validation, that is, the population was randomly divided into five parts, including four training groups and one test group, and the Pearson correlation coefficient between the predicted value and the true value was calculated. The final correlation coefficient mean obtained by repeated cross validation is shown in Table 3. Estimation of genetic variance by RRBLUP and error variance The heritability calculation results for different traits are shown in Table 4.

[0084] Table 3 Statistics on the accuracy of genomic selection for wool traits of Alpine Merino sheep with fixed effects

[0085]

[0086] Table 4 Heritability statistics of each trait under the same fixed effect condition

[0087]

[0088] 6. Construction of a genome-wide selection index estimation model

[0089] 6.1GEBV Calculation Model

[0090] The GEBV calculation model is as follows:

[0091]

[0092] Where y represents the corrected phenotypic value vector, X is the fixed effect association matrix, a represents the fixed effect vector, and Z j is the genotype vector at site j, b j is the estimated effect value of site j, and e represents the random residual vector. By measuring the marker genotype of each individual, the estimated marker effect is gradually introduced, and finally all are accumulated to obtain the individual's genomic breeding value.

[0093] Genomic estimated breeding value GEBV of individual i i The calculation is as follows:

[0094]

[0095] Select SSGBLUP and RRBLUP for genomic breeding value estimation GEBV i The calculation is as follows:

[0096] (1)SSGBLUP

[0097] SSGBLUP (Single-Step GBLUP): Based on the kinship matrix A and the genome relationship matrix G, a new matrix H is established by integration. The H matrix construction method is:

[0098]

[0099] Where A and G are A matrix and G matrix respectively, and subscripts 1 and 2 are individuals without genotype and individuals with genotype respectively.

[0100] Since G cannot be inverted when it is a singular matrix, we define G as G ω =(1-ω)G+ωA 22 , then the H inverse matrix can be transformed into:

[0101]

[0102] In the formula, ω is the weighting factor, that is, the proportion of polygenic genetic effects. After constructing the H matrix, the process of solving the mixed model equations (MME) is the same as the traditional BLUP:

[0103]

[0104] (2)RRBLUP

[0105] RRBLUP: Ridge regression best linear unbiased prediction, first calculates the effect value of each marker, then accumulates the effect values ​​to obtain the breeding value. RRBLUP assumes that the variance of the effect values ​​of all SNP loci is equal, that is, σ g1 2 =σ g2 2 =…=σ gM 2 The effect size of the SNP can be obtained by solving the following mixed model equations:

[0106]

[0107] Where M is the number of SNP sites, h 2 The chivalrous heritability of the target trait.

[0108] Seven traits, including body weight, wool length, shearing amount, wool fiber diameter, wool fiber diameter coefficient of variation, clean wool rate and clean wool volume, were identified as breeding target traits. Studies have shown that there is a strong correlation between shearing amount, clean wool rate and clean wool volume (r = 0.7), so only "clean wool volume" was selected as the target trait involved in the calculation of the genome comprehensive selection index. The genetic standard deviation data of the clean wool volume, wool length, wool fiber diameter and body weight traits of Alpine Merino sheep mainly come from Australian Merino sheep, Chinese Merino sheep, Subo Merino sheep, Aohan fine wool sheep and previous research reports of the research group. An increase in wool fiber diameter will lead to a decrease in economic value. Among the four target traits, the economic relative importance of clean wool volume accounts for the largest proportion of the total, accounting for 43.24%, followed by body weight accounting for 25.17%, wool fiber diameter accounting for 21.33% and wool length accounting for 6.19%. When compared according to the breeding target traits, the relative importance ratio of net wool weight, wool length, wool fiber diameter and body weight is 43.24:6.19:21.33:25.17, and the approximate ratio is 4:1:2:3, see Table 5.

[0109] Table 5 Marginal benefits and relative importance of target traits in Alpine Merino sheep breeding

[0110]

[0111] Note: “-” represents the breeding direction of the trait.

[0112] 6.2 Standard economic value calculation method

[0113] The marginal benefit (i.e. economic value) of each breeding target trait is calculated using the “difference method”, and the formula is as follows:

[0114] v i =∑ΔR ij -∑ΔC ij

[0115] where v i is the economic value of target trait i; ΔR ij is the jth output component of trait i; ΔC ij is the jth input component of trait i. When calculating the marginal benefit of the target trait of alpine Merino sheep breeding, the economic value of the breeding target trait is calculated based on the lambing cycle of a ewe. Combined with the biological characteristics of fine-wool sheep and the current market economic environment. The breeding goal should consider both economic factors and genetic factors. The marginal benefit (Vi) of each breeding target trait and its corresponding genetic standard deviation (σ A ) multiplied by (V σAi ) as the standard economic value.

[0116] 6.3 Genomic comprehensive selection index estimation model

[0117] The genomic breeding value (GEBV) of the four breeding target traits of fine-wool sheep, namely weight, wool length, wool fiber diameter and net wool yield, was calculated as the denominator, and the standard economic value of each trait was used as the numerator to formulate the fine-wool sheep genomic comprehensive selection index (Genomic Performance Index, GPI).

[0118] The estimation model of the comprehensive selection index of fine wool sheep genome is as follows:

[0119]

[0120] Where, GPI: genome comprehensive selection index; GEBV: CFW : Genomic Estimated Breeding Value for Gross Weight; GEBV SL : Genomic Estimated Breeding Value for Wool Length; GEBV FD : Genomic Estimated Breeding Value for Wool Fiber Diameter; GEBV WT : Genomic estimated breeding value of body weight; VσA CFW : Net gross genome standard economic value; VσA SL : Economic value of wool length genomic standard; VσA FD : Economic value of wool fiber diameter genome standard; VσA WT : Economic value of genomic standards for body weight.

[0121] The results of the analysis of prediction accuracy based on the above fixed effects and model selection showed that the breeding values ​​calculated by the RRBLUP and SSGBLUP models under the condition of adding only "group" as the fixed effect were used to construct a genomic comprehensive selection index model based on the target trait. The genomic comprehensive selection index of individuals was calculated and ranked. The results showed that among the top 100 individuals, 94 individuals were consistent and 6 individuals were different, as shown in Table 6.

[0122] Table 6 Comparison of the ranking of comprehensive selection index of candidate individual genomes

[0123]

[0124]

[0125]

[0126]

[0127] Note: Different individuals are indicated by *.

[0128] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the scope of protection of the present invention.

Claims

1. A method for genomic selection breeding of fine wool sheep, characterized in that: The method comprises the following steps: (1) Calculate the genomic estimated breeding values ​​(GEBV) of four target traits of fine wool sheep, namely weight, wool length, wool fiber diameter and wool mass; (2) Calculate the standard economic value (VσA) of the four target traits of fine wool sheep: weight, wool length, wool fiber diameter and net wool yield; (3) Constructing a comprehensive selection index estimation model for fine wool sheep genome: as shown in formula (I), Where, GPI: genome comprehensive selection index; GEBV: CFW : Genomic Estimated Breeding Value for Gross Weight; GEBV SL : Genomic Estimated Breeding Value for Wool Length; GEBV FD : Genomic Estimated Breeding Value for Wool Fiber Diameter; GEBV WT : Genomic estimated breeding value of body weight; VσA CFW : Net gross genome standard economic value; VσA SL : Economic value of wool length genomic standard; VσA FD : Economic value of wool fiber diameter genome standard; VσA WT : Economic value of the body weight genomic standard; (4) Select fine-wool sheep individuals with high GPI values ​​for breeding.

2. The method according to claim 1, characterized in that The GEBV calculation model is shown in formula (II): In the formula, GEBV i is the genomic estimated breeding value of trait i, Z j is the genotype vector at site j; b j is the estimated effect size of site j.

3. The method according to claim 1, characterized in that The VσA calculation model is shown in formula (III): v σAi =v i ×σ A ,(Ⅲ); In the formula, v σAi is the standard economic value of trait i; v i is the economic value of trait i; σ A For v i The corresponding genetic standard deviation.

4. The method according to claim 3, characterized in that The v i The difference method is used for calculation, and the formula is shown in formula (IV): in i =∑ΔR ij -∑ΔC ij ,(Ⅳ); In the formula, ΔR ij is the jth output component of trait i; ΔC ij is the jth input component of trait i; when calculating the economic value of the target traits of Alpine Merino sheep breeding, it is calculated based on the lambing cycle of one ewe.

5. The method according to claim 2, characterized in that The GEBV i Select either the SSGBLUP or RRBLUP method for calculations.

6. The method according to claim 5, characterized in that The SSGBLUP method is: Based on the kinship matrix A and the genome relationship matrix G, a new matrix H is established by integration: In formula (V), A and G are A matrix and G matrix respectively, and subscripts 1 and 2 are individuals without genotype and individuals with genotype respectively; since G cannot be inverted when it is a singular matrix, G is defined as: G ω =(1-ω)G+ωA 22 ,(Ⅵ); Then the H inverse matrix can be transformed into: In formula (VII), ω is the weighting factor, i.e., the proportion of polygenic genetic effects; After constructing the H matrix, it solves the system of equations for the mixed model:

7. The method according to claim 5, characterized in that The RRBLUP method is as follows: each marker effect value is calculated, and then the effect values ​​are accumulated to obtain the breeding value; Assume that the variance of the effect values ​​of all SNP sites is equal, that is: s g1 2 =s g2 2 =…=s gM 2 ,(Ⅸ); The effect size of the SNP can be obtained by solving the following mixed model equations: Where M is the number of SNP sites, h 2 The chivalrous heritability of the target trait.

8. The method according to claim 7, characterized in that The genetic variance was estimated using the kin.blup function in the RRBLUP package of the software. and error variance The heritability calculation formula is:

9. The method according to claim 8, characterized in that The accuracy evaluation formula of GEBV is: In the formula, y represents the corrected phenotypic value.

10. The method according to claim 9, characterized in that The y calculation formula is: Where X is the fixed effect correlation matrix, a represents the fixed effect vector, and e represents the random residual vector.

Images