A SNP locus for efficiently predicting plant height traits of cut chrysanthemum and its method and application

Abstract

The present invention discloses a highly efficient prediction SNP site of cut chrysanthemum plant height trait, and also discloses a highly efficient prediction SNP site method of cut chrysanthemum plant height trait and its application. The present invention uses a large number of SNP markers that are widely distributed, can be stably inherited and are easy to high-throughput genotyping to perform genome-wide association analysis of cut chrysanthemum plant height related traits. The results show that the best markers related to plant height traits have prediction accuracies of 0.97, 0.98, 0.97 and 0.96 for plant height, internode number, internode length and stem diameter, respectively. It can be applied to the early prediction of cut chrysanthemum plant height, internode length, internode number and stem diameter, shorten the breeding cycle, provide site information for the development of functional markers for chrysanthemum plant height related traits and genetic improvement, and have important theoretical and practical significance for the selection of excellent chrysanthemum materials with different plant types.

Description

Technical Field

[0001] The invention belongs to the field of biotechnology and relates to a SNP site for efficiently predicting plant height traits of cut chrysanthemum and a method and application thereof. Background Art

[0002] Chrysanthemum (Chrysanthemum morifolium Ramat.) is a perennial perennial flower of the genus Chrysanthemum in the family Asteraceae. It is one of the traditional famous flowers in my country and one of the four major cut flowers. It has high ornamental and economic value (Luo C et al. SSR analysis of genetic relationship and classification in chrysanthemum germplasm collection. Horticultural Plant Journal, 2018. 4 (2): 73-82). Plant height is one of the key traits that determine the plant type of chrysanthemum and is also an important criterion for the quality classification of cut chrysanthemum. The plant height of high-quality cut flower varieties is usually greater than 80 cm (Ma Wanru et al. Construction of stem and branch characteristics evaluation system and variety evaluation of multi-headed cut chrysanthemum varieties. Chinese Agricultural Science, 2019. 52 (14): 2515-2524), but too high plant height will lead to problems such as easy lodging and high energy consumption in production. Therefore, ideal plant height has always been an important goal of cut chrysanthemum breeding. The number of internodes, internode length and stem diameter are the main components of plant height and are also important indicators of the ideal plant type of cut chrysanthemum. Chrysanthemum plant height is a quantitative trait controlled by multiple genes and has a high heritability (Zhang F et al. Mapping single-locus and epistatic quantitative trait loci for plant architectural traits in chrysanthemum. Molecular Breeding, 2011. 30(2): 1027-1036; Su J et al. Combining ability, heterosis, genetic distance and their intercorrelations for waterlogging tolerance traits inchrysanthemum. Euphytica, 2017. 213(2): e42). In addition, there is a significant phenotypic correlation between chrysanthemum plant height and traits such as the number of internodes, internode length, stem diameter, flower diameter, and flowering period (Han Yong et al. Correlation and path analysis of standard cut chrysanthemum diameter and other important quantitative traits. Journal of Nanjing Agricultural University, 2012. 35(1): 33-37.).

[0003] Genome wide association study (GWAS) is an effective method to locate genetic loci associated with target traits by detecting the degree of association between genome-wide genetic markers and phenotypic variations.At present, GWAS analysis has been used to identify genes related to chrysanthemum plant type (Li P et al. Genetic diversity, population structure and association analysis in cut chrysanthemum (Chrysanthemummorifolium Ramat.). Molecular Genetics and Genomics, 2016. 291(3): 1117-1125), flower type (Chong X et al. Identification of favorable SNP alleles and candidate genes responsible for inflorescence-related traits via GWAS in chrysanthemum. Plant Molecular Biology, 2019. 99(4-5): 407-420), aphid resistance (Fu X et al. Genetic variation and association mapping of aphid (Macrosiphoniella sanbourni) resistance in chrysanthemum (Chrysanthemum morifolium Ramat.). Euphytica, 2018. 214(2): e21), drought tolerance (Li P et al. al. Association analysis of drought tolerance in cutchrysanthemum(Chrysanthemummorifolium Ramat.)at seedling stage.3Biotech,2018.8(5):e226) and waterlogging tolerance (Su J,et al. Genome-wide association study identifies favorable SNP alleles and candidate genes for waterlogging tolerance inchrysanthemums.Horticulture Research,2019.6:e21) and other favorable variant sites related to traits. However, since most of the important horticultural traits of chrysanthemum are quantitative traits and are controlled by multiple micro-effect genes at the same time, and most of the identified sites are environment-specific sites, they have not yet been applied in actual breeding processes.Unlike whole-genome association analysis, whole-genome selection (GS) does not require the pre-determination of molecular markers with significant effects, but predicts the estimated breeding value of individuals by evaluating the genetic effects of all markers in the genome, especially for complex traits controlled by micro-effect polygenes (Juliana P et al. Improving grain yield, stress resistance and quality of bread wheat using large-scale genomics. Nature Genetics, 2019. 51 (10): 1530-1539). With the rapid development of high-throughput sequencing technology, more and more studies have shown that incorporating association sites identified by GWAS into GS models can improve the predictive ability of GS to a certain extent. This method has achieved certain results in animal and plant breeding, but it has not yet been reported in chrysanthemum plant height-related traits. Summary of the invention

[0004] Purpose of the invention: In view of the defects and shortcomings of the prior art, the present invention provides a SNP site for efficiently predicting the plant height trait of cut chrysanthemum. The prediction accuracy of this site for plant height, number of internodes, internode length and stem diameter reaches 0.97, 0.98, 0.97 and 0.96 respectively, which can be used to predict traits related to plant height of cut chrysanthemum, shorten the breeding cycle and accelerate the breeding process.

[0005] The invention provides a method for efficiently predicting SNP sites of plant height traits of cut chrysanthemum.

[0006] The present invention also provides an efficient prediction method of a SNP site for the plant height trait of cut chrysanthemum and its application in predicting the plant height of cut chrysanthemum.

[0007] Technical solution: In order to achieve the above-mentioned purpose, the present invention provides a method for efficiently predicting the SNP loci of the plant height trait of cut chrysanthemum. The SNP loci of the cut chrysanthemum trait and their specific nucleotides are as follows:

[0008]

[0009]

[0010] The method for efficiently predicting the SNP loci of the plant height trait of cut chrysanthemum of the present invention comprises the following steps:

[0011] Step 1: Select cut chrysanthemum varieties from different sources and without direct genetic relationship, and count the plant height-related traits of the population under different environments at the end of the growth period;

[0012] Step 2: Perform descriptive statistics and variance analysis on the cut chrysanthemum traits counted in step 1, and calculate the broad-sense heritability of each trait;

[0013] Step 3: GBS (genotyping-by-sequencing) was used to genotype the cut chrysanthemum varieties in step 1, and the clean reads were aligned to the chrysanthemum reference genome using BWA software. SAMTOOLS software was used for SNP calling, and VCFtools software was used to filter the detected SNPs. The obtained high-quality SNPs were used for genome-wide association analysis;

[0014] Step 4: Based on the phenotypic data obtained in step 1 and the genotypic data obtained in step 3, a genome-wide association analysis was performed using the EMMAX method, and SNPs significantly associated with each trait were screened according to the significance threshold;

[0015] Step 5: Conduct genome-wide selection analysis on cut chrysanthemum traits, determine the best marker set by comparing the prediction accuracy of different SNP marker sets, and screen materials that meet breeding value requirements.

[0016] Furthermore, the cut chrysanthemum plant height-related trait is any one of cut chrysanthemum plant height, internode number, internode length or stem thickness.

[0017] Furthermore, the method for statistically analyzing cut chrysanthemum traits in step 1 is to select plants of each variety with consistent growth under different environments for phenotypic measurement.

[0018] Preferably, the number of environments in step 1 is 2, environment one is Nanjing Hushu E1, and environment two is Nanjing Suoshi E2. In mid-October 2020, 5 plants with consistent growth are selected for each variety for phenotypic measurement.

[0019] Furthermore, the phenotypic data analysis of cut chrysanthemum in step 2 was mainly completed using R language, and the variance analysis model was Y ij =μ+G i +Yj+GY ij +e ij .

[0020] Preferably, the variance analysis model Y ij =μ+G i +Yj+GY ij +e ij Middle Y ij is the phenotypic observation value, μ is the mean under multiple environments, G i is the genotype effect, Yj is the environmental effect, GY ij is the genotype-environment interaction effect, e ijis a random error. The broad heritability (H) of each trait was calculated based on the results of variance analysis. 2 ), whose expression is H 2 =σ2g / (σ2g+σ2ge / n+σ2e / rn), σ2g is the genotype variance, σ2ge is the genotype-environment interaction variance, σ2e is the residual variance, n is the number of environments, and r is the number of replications of a single environment.

[0021] Furthermore, in step 3, SNPs were filtered using VCFtools software according to the standard of “--minDP 6--min-alleles2--max-alleles 2--max-missing 0.85--maf 0.05”.

[0022] Furthermore, in step 4, the EMMAX method takes the principal component analysis (PCA) and the kinship K matrix into the model as covariates, uses the EMMAX software to perform genome-wide association analysis on the cut chrysanthemum traits under different environments, calculates the significance of each SNP site, i.e., the P value, and sets P < 1 × 10 -5 Potential candidate SNPs were screened out as the threshold for significant association.

[0023] Furthermore, the step 5 specifically includes the following steps:

[0024] (1) Since the phenotypic data of some varieties are missing, they need to be filled in during the whole genome selection analysis; if the phenotypic data are missing, the impute() function in the Hmisc package of the R software is used for automatic interpolation; if the genotype data are missing, the A.mat() function in the rrBLUP package is used for automatic interpolation, and the genotype format of the SNP is encoded as the {-1,0,1} genotype format;

[0025] (2) Setting different labeling schemes to predict target traits under different environments;

[0026] (3) The five-fold cross-validation method was used to perform whole-genome selection using the rrBLUP model. The correlation coefficient between the genomic estimated breeding value (GEBV) and the actual observed value was used as an indicator to evaluate the prediction accuracy. The marker set with the largest correlation coefficient was the optimal SNP marker set.

[0027] Preferably, the expression of the rrBLUP model is: y=WGu+ε, y is the phenotype vector, u~N(0,Iσ 2 u ) is the marker effect vector, W represents the association matrix between genotype and phenotype, G represents the genetic matrix, ε is the random error, and the solution of the model is: Where Z = WG, λ = σ 2e / σ 2 u It refers to the ratio of the residual variance to the marker effect variance.

[0028] Furthermore, the different marker schemes are specifically as follows: sorting the SNPs from small to large according to the P value calculated by the EMMAX model, selecting the SNPs below the significant threshold and the top 100 to 2000 SNPs significantly associated with the target trait as scheme one, or randomly selecting the same number of SNPs as the scheme from the high-quality SNPs obtained in the above step 3 as scheme two; using these two schemes to predict the target traits under different environments, respectively, and determine the optimal SNP marker set.

[0029] The invention discloses an application of the SNP site for efficiently predicting the height of cut chrysanthemum plants in predicting the height of cut chrysanthemum plants.

[0030] The method for efficiently predicting the plant height trait of cut chrysanthemum according to the present invention is applied in predicting the plant height of cut chrysanthemum.

[0031] Further, the application of the SNP loci for efficiently predicting traits related to plant height of cut chrysanthemum or the method for efficiently predicting SNP loci for traits related to plant height of cut chrysanthemum in predicting plant height of cut chrysanthemum comprises the following steps:

[0032] (1) Based on the best SNP marker set screened, the formula Y = μ + Xg + e was used to estimate the GEBV (genomic estimated breeding value) of plant height for 200 different cut chrysanthemum varieties;

[0033] (2) When the breeding value of the plant height of a cut chrysanthemum variety is greater than 80 cm, the variety can be used as a high-quality cut chrysanthemum variety.

[0034] Based on the existing germplasm resource population, the present invention uses GWAS (genome-wide selection association analysis) to screen SNP markers significantly associated with plant height-related traits, providing site information for genetic improvement of chrysanthemum plant type. In addition, based on GWAS, molecular markers related to target traits are used to estimate the breeding value of individuals through GS (genome-wide selection), so as to achieve early prediction and selection of plant height traits of cut chrysanthemum, thereby shortening the breeding cycle and accelerating the breeding process.

[0035] Beneficial effects of the present invention: Compared with the prior art, the present invention has the following advantages:

[0036] (1) The present invention uses SNP markers that are numerous, widely distributed, stably inherited, and easy to high-throughput genotype to perform whole-genome association analysis of traits related to plant height of cut chrysanthemum. Through whole-genome association analysis, SNP markers related to four traits, namely plant height, number of internodes, internode length, and stem thickness, are identified, providing site information for the development of functional markers for traits related to plant height of chrysanthemum and genetic improvement.

[0037] (2) Based on the whole genome association analysis, the present invention combines the identified SNP markers for whole genome selection, and screens out the best SNP markers that can effectively predict the plant height-related traits of cut chrysanthemum under different SNP marker sets. The prediction accuracies of these best markers for plant height, number of internodes, internode length and stem diameter are 0.97, 0.98, 0.97 and 0.96, respectively. It is preliminarily believed that they can be applied to the early prediction of plant height-related traits of cut chrysanthemum, facilitate the rapid screening of excellent cut chrysanthemum materials with ideal plant height, shorten the breeding cycle, and have important theoretical and practical significance for breeding new varieties of cut chrysanthemum with ideal plant types. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 The phenotypic distribution histograms of cut chrysanthemum in environment E1 and environment E2, where A is plant height, B is internode number, C is internode length, and D is stem diameter;

[0039] Figure 2 This is the genome-wide prediction accuracy of cut chrysanthemum plant height under random selection and trait-related SNPs, where Figure A is the E1 environment and Figure B is the E2 environment.

[0040] Figure 3 This is the genome-wide prediction accuracy of the number of internodes in cut chrysanthemum under random selection and trait-related SNPs, where Figure A is the E1 environment and Figure B is the E2 environment.

[0041] Figure 4 This is the genome-wide prediction accuracy of internode length of cut chrysanthemum under random selection and trait-related SNPs, where Figure A is the E1 environment and Figure B is the E2 environment.

[0042] Figure 5 This is the genome-wide prediction accuracy of cut chrysanthemum stem thickness under random selection and trait-related SNPs, where Figure A is the E1 environment and Figure B is the E2 environment.

[0043] Figure 6 It is the genome-wide prediction accuracy of four traits: plant height, number of internodes, internode length, and stem diameter based on the optimal SNP marker set and the means of the two environmental phenotypes.

[0044] Figure 7 Flow chart of a method for efficiently predicting traits related to plant height of cut chrysanthemum based on whole genome selection. DETAILED DESCRIPTION

[0045] The present invention is further described in detail below in conjunction with the embodiments.

[0046] Example 1

[0047] 1. Experimental materials and field design

[0048] 200 cut chrysanthemum varieties from different sources and without direct genetic relationship were selected as test materials (Table 1). All materials came from the China Chrysanthemum Germplasm Conservation Center of Nanjing Agricultural University. Those skilled in the art can obtain the above germplasm from the "China Chrysanthemum Germplasm Conservation Center". They were planted in the Hushu Base (E1) and Suoshi Base (E2) of Nanjing Agricultural University in July 2020. The row spacing and plant spacing of the plot were both 0.10m, and water, fertilizer, pesticide and other field management technical measures were operated according to the high-quality production technology of chrysanthemum. At the end of vegetative growth (late October), 5 plants with consistent growth of each variety were selected to measure plant height, number of internodes, internode length and stem diameter. Measurement method: Plant height (PH / cm): height from stem base to top of corolla; number of internodes (IN / ): number of all internodes from stem base to top of corolla; internode length (IL / cm): average length of internodes in the middle and upper part of the plant; stem diameter (SD / mm): diameter of stem in the middle and upper part of the plant.

[0049] Table 1 200 cut chrysanthemum varieties tested

[0050]

[0051]

[0052] (II) Statistical analysis of phenotypic data

[0053] The phenotypic distribution of plant height, number of internodes, internode length and stem diameter of 200 cut chrysanthemum varieties under E1 and E2 environments is shown in the figure. Figure 1 As shown in the figure, the R language'pastecs' package was used to perform descriptive statistics on the four traits of plant height, number of internodes, internode length and stem diameter under the E1 and E2 environments in step (I), and the 'lm4' package was used for variance analysis. The variance analysis model was Y ij =μ+G i +Yj+GY ij +e ij , where Y ij is the phenotypic observation value, μ is the mean under multiple environments, G i is the genotype effect, Yj is the environmental effect, GY ij is the genotype-environment interaction effect, e ij The broad-sense heritability (H) of each trait was calculated based on the results of variance analysis. 2 ), whose expression is H 2 =σ2g / (σ2g+σ2ge / n+σ2e / rn), σ2g is the genotype variance, σ2ge is the genotype-environment interaction variance, σ2e is the residual variance, n is the number of environments, and r is the number of replications of a single environment.

[0054] The results showed that there was a wide variation in the four traits of plant height, number of internodes, internode length and stem diameter of 200 cut chrysanthemum materials, with a coefficient of variation of 16.49-30.18% (Table 2). The broad heritability of plant height, number of internodes, internode length and stem diameter was 79.61%, 86.06%, 84.42% and 79.21%, respectively (Table 2), indicating that the four traits of plant height, number of internodes, internode length and stem diameter of cut chrysanthemum were mainly affected by genetic factors. The results of variance analysis showed that in addition to the genotype effect, the environmental effect and the genotype-environment interaction effect also had a significant effect on the traits related to plant height (Table 3).

[0055] Table 2 Descriptive statistics of plant height, number of internodes, internode length and stem diameter

[0056]

[0057] Table 3 Analysis of variance of four traits: plant height, number of internodes, internode length and stem diameter

[0058]

[0059] Note: *** indicates P < 0.001.

[0060] (III) Genotyping

[0061] GBS (genotyping-by-sequencing) genome sequencing was performed using the paired-end 150bp of the Illumina Hiseq platform, and clean reads were obtained after removing the adapters and low-quality reads that may be contained in the raw reads. The clean reads were aligned to the chrysanthemum reference genome (SongA et al. Analyses of a chromosome-scale genome assembly reveal the origin and evolution of cultivated chrysanthemum. Nat Commun, 2023.14:2021) using BWA software (parameters: mem-t 4-k 32-M), and SNP calling was performed using SAMTOOLS software. GBS sequencing and SNP marker development were completed by Beijing Novogene Bioinformatics Technology Co., Ltd. The original SNPs obtained were filtered using VCFtools software according to the criteria of "--minDP 6--min-alleles 2--max-alleles 2--max-missing0.85--maf 0.05", and finally 330,710 high-quality SNPs were obtained for further genome-wide association analysis.

[0062] (IV) Genome-wide association analysis

[0063] Based on the phenotypic data obtained in step (a) and the genotypic data obtained in step (c), a genome-wide association analysis was performed, and the EMMAX method was used to perform genome-wide association analysis on plant height, number of internodes, internode length, and stem thickness under E1 and E2, respectively. This method incorporates principal component analysis (PCA) and kinship K matrix as covariates into the model. The "emmax-kin" subroutine of EMMAX software was used to calculate the kinship K matrix between the populations. Principal component analysis was performed using GCTA software, and the first 10 principal components were selected as the covariance matrix for association analysis. The significance of each SNP locus, i.e., the P value, was calculated using EMMAX software, and P<1×10 -5 Potential candidate SNPs were screened out as the threshold for significant association (Table 4).

[0064] As shown in Table 4, a total of 12 SNP loci associated with plant height were detected, of which 3 SNPs were from E1 and the remaining 9 SNPs were from E2. Except for the two loci Chr16__103728540and Chr21__9532188, the other loci showed an enhancement effect. A total of 12 SNP loci associated with internode number were detected, of which only 2 SNPs were from E2. A total of 10 SNP loci associated with internode length were detected, of which 5 SNPs were from E1 and 5 SNPs were from E2, and the phenotypic effect values ​​ranged from 0.2 to 0.45. A total of 8 SNP loci associated with stem diameter were detected, of which only Chr8__245595683 showed a reduction effect. However, the same loci were not detected for the same trait in the two environments, indicating that environmental effects and genotype-environment interaction effects play an important role in the genetic variation of individuals. In addition, plant height and internode number shared two identical association loci, Chr12__270808434 and Chr12__270808478, and both showed synergistic effects, indicating that these traits may be controlled by pleiotropic genes or closely linked gene loci.

[0065] Table 4 SNP loci significantly associated with plant height, internode length, internode number and stem diameter

[0066]

[0067]

[0068] Note: a The bold SNP sites are alleles that were repeatedly detected; b is the phenotypic effect value; ** and * indicate significant differences at the 0.01 and 0.05 levels, respectively; ns indicates no significant difference.

[0069] 5. Whole-genome selection

[0070] Since both phenotypic and genotypic data are missing, the missing phenotypic data are automatically interpolated using the impute() function in the Hmisc package of the R software, and the missing genotypic data are automatically filled using the A.mat() function in the rrBLUP package, and the genotype format of the SNP is encoded as the {-1,0,1} genotype format. Two marker schemes were set up, and the 330,710 SNPs in step (III) were sorted from small to large according to the P value calculated by the EMMAX model, and the SNPs under the significant threshold and the top 100, 300, 500, 1000, 5000, 10000, 15000 and 20000 SNPs significantly associated with the target traits (plant height, number of internodes, internode length and stem thickness) were selected as scheme 1; from the 330,710 SNPs obtained in step (III), the same number of SNPs as the scheme were randomly selected as scheme 2; these two schemes were used to predict the target traits in the E1 and E2 environments respectively, and the best SNP marker set was determined. The phenotypic mean under E1 and E2 environments was taken as the true value, and the union of the best SNP markers under the two environments was taken as the marker set. The prediction accuracy of plant height, internode number, internode length and stem thickness was verified by whole genome selection. Whole genome selection was completed using the "rrBLUP" package of R language. The expression of the model is: y = WGu + ε, y is the phenotypic vector, u ~ N (0, Iσ 2 u ) is the marker effect vector, W represents the association matrix between genotype and phenotype, G represents the genetic matrix, ε is the random error, and the solution of the model is: Where Z = WG, λ = σ 2 e / σ 2 u Refers to the ratio of the residual variance to the marker effect variance. The five-fold cross validation method was repeated 500 times to improve the generalization ability of the model. The rrBLUP model was used for whole genome selection. The correlation coefficient between the genomic estimated breeding value (GEBV) and the actual observed value was used as an indicator to evaluate the prediction accuracy. The marker set with the largest correlation coefficient was the best SNP marker set.

[0071] The results of whole genome selection showed that the prediction accuracy of trait-related SNPs was much higher than that of randomly selected SNPs under the same number of conditions, whether in E1 or E2. For the four traits of plant height, number of internodes, internode length, and stem thickness, the prediction accuracy showed a significant increasing trend as the number of SNP markers increased to 1,000, but when the number of SNP markers increased from 1,000 to 20,000, the change trend of prediction accuracy was almost horizontal (Table 5, Figure 2-Figure 5 ). For plant height, the prediction accuracy of the most significantly correlated 1000 SNPs in E1 and E2 environments was 0.97 and 0.96 respectively; for internode number, the prediction accuracy of the most significantly correlated 1000 SNPs in E1 and E2 environments was 0.97; for internode length, the prediction accuracy of the most significantly correlated 1000 SNPs in E1 and E2 environments was 0.96; for stem diameter, the prediction accuracy of the most significantly correlated 1000 SNPs in E1 and E2 environments was 0.95 and 0.96 respectively. It can be seen that the most significantly correlated 1000 SNPs can achieve a fairly high prediction accuracy with relatively few markers, and it is determined as the optimal SNP marker set. Based on the union of the best SNP markers in the two environments and the mean of the trait phenotypes in the two environments, the prediction accuracy of plant height, internode number, internode length and stem diameter was 0.97, 0.98, 0.97, 0.96 ( Figure 6 ), further illustrating the feasibility of the SNP markers developed in the present invention for predicting plant height, internode number, internode length and stem thickness of cut chrysanthemum.

[0072] Table 5 Genome-wide prediction accuracy of plant height-related traits using different molecular marker sets

[0073]

[0074]

[0075] Note: Trait correlation_sig indicates SNPs under the significant threshold (plant height: sig = 12; number of internodes: sig = 12; internode length: sig = 10; stem thickness: sig = 8); Random selection_ran indicates the randomly selected SNPs with the same number as the significant threshold (plant height: ran = 12; number of internodes: ran = 12; internode length: ran = 10; stem thickness: ran = 8).

[0076] Example 2 Selecting excellent breeding materials based on the best SNP marker set screened

[0077] According to the optimal SNP marker set (Table 5) selected in Example 1 and the flowchart of the whole gene selection prediction ( Figure 7 ), the breeding values ​​(GEBV) of plant height of 200 different cut chrysanthemum varieties were estimated using the formula Y = μ + Xg + e, where Y is the predicted phenotypic value, μ is the phenotypic mean of the training population, X is the phenotypic mean of the test population, g is the marker effect matrix, and e is the random effect matrix. Varieties with breeding values ​​(predicted phenotypic values) greater than 80 cm were selected as the preferred varieties of cut chrysanthemum.

[0078] Table 6 High-quality varieties with plant heights that meet the production standards for cut chrysanthemums based on the optimal GS SNP marker set

[0079]

[0080] As shown in Table 6, there are 40 varieties with breeding values ​​of plant height greater than 80 cm, which can be used as candidate materials for high-quality cut chrysanthemum and used in subsequent plant type breeding.

Claims

1. A method for efficiently predicting traits related to plant height of cut chrysanthemum, characterized in that: The following steps are involved: Step 1: Select cut chrysanthemum varieties from different sources and without direct genetic relationship, and count the plant height-related traits of the population under different environments at the end of the growth period; Step 2: Perform descriptive statistics and variance analysis on the cut chrysanthemum traits counted in step 1, and calculate the broad-sense heritability of each trait; Step 3: GBS (genotyping-by-sequencing) was used to genotype the cut chrysanthemum varieties in step 1, and the cleanreads were aligned to the chrysanthemum reference genome using BWA software. SAMTOOLS software was used for SNP calling, and VCFtools software was used to filter the detected SNPs. The obtained high-quality SNPs were used for genome-wide association analysis; Step 4: Perform genome-wide association analysis based on the phenotypic data obtained in step 1 and the genotypic data obtained in step 3, and screen SNPs that are significantly associated with each trait based on the significance threshold; Step 5: Conduct genome-wide selection analysis on traits related to plant height of cut chrysanthemum, determine the best marker set by comparing the prediction accuracy of different SNP marker sets, and select materials that meet the breeding value requirements; Wherein, the cut chrysanthemum plant height-related trait in step 1 is any one of the cut chrysanthemum plant height, number of internodes, internode length or stem thickness; The variance analysis model for the cut chrysanthemum phenotype data analysis in step 2 is Y ij =μ+G i +Yj+GY ij +e ij , where Y ij is the phenotypic observation value, μ is the mean under multiple environments, G i is the genotype effect, Yj is the environmental effect, GY ij is the genotype-environment interaction effect, e ij The broad-sense heritability (H) of each trait was calculated based on the results of variance analysis. 2 ), whose expression is H 2 =σ2g / (σ2g+σ2ge / n+σ2e / rn), σ2g is the genotype variance, σ2ge is the genotype-environment interaction variance, σ2e is the residual variance, n is the number of environments, and r is the number of replicates for a single environment; In step 4, genome-wide association analysis was performed on the cut chrysanthemum traits under different environments, and the significance of each SNP site, i.e., the P value, was calculated. -5 Potential candidate SNPs were screened as the threshold for significant association; In step 5, according to the selected optimal SNP marker set, the breeding value GEBV of plant height of each different cut chrysanthemum variety is estimated using the formula Y=μ+Xg+e; wherein Y is the predicted phenotypic value, μ is the phenotypic mean of the training population, X is the phenotypic mean of the test population, g is the marker effect matrix, and e is the random effect matrix; when the breeding value of the plant height of the cut chrysanthemum variety is greater than 80 cm, the variety can be used as a cut chrysanthemum variety with high-quality plant height-related traits.

2. The method for efficiently predicting traits related to plant height of cut chrysanthemum according to claim 1, characterized in that: In step 3, SNPs were filtered using VCFtools software according to the standard of "--minDP6--min-alleles2--max-alleles2--max-missing0.85--maf0.05".

3. The method for efficiently predicting traits related to plant height of cut chrysanthemum according to claim 1, characterized in that: The step 5 specifically includes the following steps: (1) If the phenotypic data are missing, the impute() function in the Hmisc package of R software is used for automatic interpolation. If the genotypic data are missing, the A.mat() function in the rrBLUP package is used for automatic interpolation, and the genotype format of the SNP is encoded as the {-1,0,1} genotype format; (2) Setting different labeling schemes to predict target traits under different environments; (3) The five-fold cross-validation method was used to perform whole-genome selection using the rrBLUP model. The correlation coefficient between the genomic estimated breeding value (GEBV) and the actual observed value was used as an indicator to evaluate the prediction accuracy. The marker set with the largest correlation coefficient was the optimal SNP marker set.

4. An application of the method for efficiently predicting traits related to plant height of cut chrysanthemum according to any one of claims 1 to 3 in predicting traits related to plant height of cut chrysanthemum, wherein the trait related to plant height of cut chrysanthemum is any one of plant height, number of internodes, internode length or stem thickness of cut chrysanthemum.

Images