A method for whole genome selection breeding based on representative populations

Abstract

The present invention discloses a method for whole genome selection breeding based on representative populations, and belongs to the field of biological breeding technology. The present invention constructs an evolutionary tree of a reference population using molecular markers. Similar populations will be at a node on the evolutionary tree, and 3-5 individuals are extracted at the node to represent all individuals on this node to form a population 1. Then, three samples of high, medium and low are extracted according to the size of the phenotypic value of the reference population to form a population 2. After merging populations 1 and 2, the RHBLUP algorithm is used for calculation. The RHBLUP algorithm of the present invention first estimates the genetic variance of the reference population using the random HE method, and then calculates the molecular marker effect value using the BLUP method. The prediction accuracy obtained by the calculation method is relatively high. Therefore, the method of the present invention can not only significantly shorten the calculation time when processing large sample data, but also ensure the prediction accuracy at low sample sizes.

Description

Technical Field

[0001] The present invention relates to the field of biological breeding technology, and in particular to a method for whole genome selection breeding based on representative populations. Background Art

[0002] Most important agronomic traits such as yield, disease resistance and drought resistance are quantitative traits controlled by small polygenes. Traditional molecular marker-assisted breeding (MAS) technology is difficult to genetically improve such traits. The emergence of whole genome selection technology has brought breakthrough progress in the improvement of such traits. This technology is currently the most effective molecular breeding technology for quantitative trait improvement. The basic idea of ​​whole genome selection is: (1) obtain genotypes by sequencing the reference population; (2) perform phenotypic measurement on the target traits of the reference population; (3) use the whole genome selection algorithm to calculate the effect value of each marker; (4) calculate the breeding value of each individual in the test population with only genotypes; (5) select the best individual in the test population based on the breeding value. With the development of high-throughput sequencing and phenotyping technology, sample data can reach hundreds of thousands or even tens of millions. The sample limit of the most popular whole genome selection algorithm GBLUP is 150,000. Therefore, it is difficult for existing algorithms to effectively process large sample data. Summary of the invention

[0003] In order to solve the above-mentioned deficiencies in the prior art, the purpose of the present invention is to provide a method for whole genome selection breeding based on representative populations, so as to solve the problem that the existing whole genome selection algorithm has a long calculation time and cannot process large sample data.

[0004] The technical solution of the present invention to solve the above technical problems is as follows: a method for whole genome selection breeding based on a representative population is provided, comprising the following steps:

[0005] S1, selecting a preset number of molecular markers in the reference population to construct an evolutionary tree, selecting multiple individuals at each node of the evolutionary tree, and merging the individuals extracted at all nodes to obtain population 1;

[0006] S2, sorting the individuals from high to low according to the phenotypic values ​​of the target traits of the reference population, and then extracting multiple individuals with the highest, middle and lowest phenotypic values ​​respectively, merging these individuals to obtain population 2; among them, the number of individuals with the highest, middle and lowest phenotypic values ​​is the same, and each accounts for 15-20% of the number of individuals in the reference population;

[0007] S3, merging population 1 and population 2 as the representative population, and using the whole genome selection algorithm RHBLUP algorithm to calculate the marker effect value of the representative population;

[0008] S4. Calculate the breeding value of each individual in the test population according to the marker effect value, and select the best individual according to the breeding value.

[0009] Furthermore, in step S1, 300-600 molecular markers are randomly selected from the reference population to construct an evolutionary tree of the reference population, 3-5 individuals are randomly selected at each node of the evolutionary tree, and then merged to obtain population 1.

[0010] Furthermore, step S3 uses the whole genome selection algorithm RHBLUP algorithm to calculate the representative population marker effect value, including: using the random HE method to estimate the genetic variance of the reference population; and using the BLUP method to calculate the molecular marker effect value based on the genetic variance of the reference population.

[0011] Furthermore, the formula for estimating the genetic variance of the reference population using the random HE method is: Where tr represents the trace of the matrix; M represents the genetic correlation matrix; S represents the trace of the genetic relationship matrix; n represents the reference population size; represents genetic variance; represents the residual variance; y represents the phenotypic value of the reference population; T represents transpose.

[0012] Furthermore, the formula for calculating the molecular marker effect value using the BLUP method is: Among them, u represents the marker effect value; represents genetic variance; U represents the reference population marker matrix; T represents transposition; V -1 represents the inverse matrix of the sample variance of the reference population; y represents the phenotypic value of the reference population; A represents the fixed effect matrix.

[0013] Further, the calculation formula of the breeding value in step S4 is: C = U c u; where C represents the individual breeding value of the test population, U c represents the marker matrix of the test population, and u represents the marker effect value.

[0014] The present invention has the following beneficial effects: in order to obtain a representative population, an evolutionary tree of a reference population is constructed using molecular markers, and similar populations and difference populations are distinguished by the evolutionary tree. Similar populations will be at one node on the evolutionary tree, and extracting 3-5 individuals at the node can well represent all individuals at this node, forming population 1. Then, three samples of high, medium and low are extracted according to the size of the phenotypic value of the reference population to form population 2. The method of extracting samples using the present invention can greatly reduce the number of samples in the reference population. After merging population 1 and population 2, the RHBLUP algorithm is used for calculation. The RHBLUP algorithm of the present invention first uses the random HE method to estimate the genetic variance of the reference population, and then uses the BLUP method to calculate the molecular marker effect value. The prediction accuracy obtained by this calculation method is relatively high. Therefore, the method of the present invention can not only significantly shorten the calculation time when processing large sample data, but also ensure the prediction accuracy of low sample size. DETAILED DESCRIPTION

[0015] The principles and features of the present invention are described below in conjunction with the examples, which are only used to explain the present invention and are not intended to limit the scope of the present invention. If no specific conditions are specified in the examples, the conditions are carried out according to conventional conditions or the conditions recommended by the manufacturer. If the manufacturer of the reagents or instruments is not specified, they are all conventional products that can be purchased commercially.

[0016] Example 1

[0017] A method for whole genome selection breeding based on a representative population, comprising the following steps:

[0018] (1) Randomly select 500 molecular markers from the reference population to construct an evolutionary tree of the reference population. Randomly select 3 individuals at each node of the evolutionary tree and then merge them to obtain population 1.

[0019] (2) Sort the individuals according to the phenotypic values ​​of the target trait of the reference population from high to low, and then extract multiple individuals with the highest, middle, and lowest phenotypic values, respectively, and merge these individuals to obtain population 2; among them, the number of individuals with the highest, middle, and lowest phenotypic values ​​is the same, and each accounts for 15% of the number of individuals in the reference population (if the number of individuals with the highest, middle, and lowest phenotypic values ​​extracted according to the ratio of 15% is not an integer, the number of individuals extracted is determined according to the principle of rounding off);

[0020] (3) merge groups 1 and 2 to obtain a representative group;

[0021] (4) Then, the whole genome selection algorithm RHBLUP is used to calculate the representative population obtained in step (3), first using the random HE method to estimate the genetic variance of the reference population, and then using the BLUP method to obtain all marker effect values;

[0022] Among them, the formula for estimating the genetic variance of the reference population using the random HE method is as follows:

[0023]

[0024] Here tr represents the trace of the matrix; M represents the genetic correlation matrix; S represents the trace of the genetic relationship matrix; n represents the reference population size; represents genetic variance; represents the residual variance; y represents the phenotypic value of the reference population; T represents transpose.

[0025] The formula for calculating the molecular marker effect value using the RHBLUP method is as follows:

[0026]

[0027] Here u represents the marker effect value; represents genetic variance; U represents the reference population marker matrix; T represents transposition; V -1 represents the inverse matrix of the sample variance of the reference population; y represents the phenotypic value of the reference population; A represents the fixed effect matrix;

[0028] (5) using the marker effect value obtained in step (4) to calculate the breeding value of each individual in the test population, and then sorting the breeding values ​​of each individual in the test population to select the best individual;

[0029] The breeding value calculation formula is as follows:

[0030] C=U c u

[0031] Here C represents the individual breeding value of the test population, U c represents the marker matrix of the test population, and u represents the marker effect value.

[0032] Example 2

[0033] A method for whole genome selection breeding based on a representative population, comprising the following steps:

[0034] (1) 300 molecular markers were randomly selected from the reference population to construct an evolutionary tree of the reference population. Five individuals were randomly selected at each node of the evolutionary tree and then merged to obtain population 1.

[0035] (2) Sort the individuals according to the phenotypic values ​​of the target trait of the reference population from high to low, then extract multiple individuals with the highest, middle, and lowest phenotypic values, respectively, and merge these individuals to obtain population 2; among them, the number of individuals with the highest, middle, and lowest phenotypic values ​​is the same, and each accounts for 20% of the number of individuals in the reference population (if the number of individuals with the highest, middle, and lowest phenotypic values ​​extracted according to the ratio of 20% is not an integer, the number of individuals extracted is determined according to the principle of rounding off);

[0036] (3) merge groups 1 and 2 to obtain a representative group;

[0037] (4) Then, the whole genome selection algorithm RHBLUP is used to calculate the representative population obtained in step (3), first using the random HE method to estimate the genetic variance of the reference population, and then using the BLUP method to obtain all marker effect values;

[0038] Among them, the formula for estimating the genetic variance of the reference population using the random HE method is as follows:

[0039]

[0040] Here tr represents the trace of the matrix; M represents the genetic correlation matrix; S represents the trace of the genetic relationship matrix; n represents the reference population size; represents genetic variance; represents the residual variance; y represents the phenotypic value of the reference population; T represents transpose.

[0041] The formula for calculating the molecular marker effect value using the RHBLUP method is as follows:

[0042]

[0043] Here u represents the marker effect value; represents genetic variance; U represents the reference population marker matrix; T represents transposition; V -1 represents the inverse matrix of the sample variance of the reference population; y represents the phenotypic value of the reference population; A represents the fixed effect matrix;

[0044] (5) using the marker effect value obtained in step (4) to calculate the breeding value of each individual in the test population, and then sorting the breeding values ​​of each individual in the test population to select the best individual;

[0045] The breeding value calculation formula is as follows:

[0046] C=U c u

[0047] Here C represents the individual breeding value of the test population, U c represents the marker matrix of the test population, and u represents the marker effect value.

[0048] Example 3

[0049] A method for whole genome selection breeding based on a representative population, comprising the following steps:

[0050] (1) 600 molecular markers were randomly selected from the reference population to construct an evolutionary tree of the reference population. Three individuals were randomly selected at each node of the evolutionary tree and then merged to obtain population 1.

[0051] (2) Sort the individuals according to the phenotypic values ​​of the target trait of the reference population from high to low, and then extract multiple individuals with the highest, middle, and lowest phenotypic values, respectively, and merge these individuals to obtain population 2; among them, the number of individuals with the highest, middle, and lowest phenotypic values ​​is the same, and each accounts for 15% of the number of individuals in the reference population (if the number of individuals with the highest, middle, and lowest phenotypic values ​​extracted according to the ratio of 15% is not an integer, the number of individuals extracted is determined according to the principle of rounding off);

[0052] (3) merge groups 1 and 2 to obtain a representative group;

[0053] (4) Then, the whole genome selection algorithm RHBLUP is used to calculate the representative population obtained in step (3), first using the random HE method to estimate the genetic variance of the reference population, and then using the BLUP method to obtain all marker effect values;

[0054] Among them, the formula for estimating the genetic variance of the reference population using the random HE method is as follows:

[0055]

[0056] Here tr represents the trace of the matrix; M represents the genetic correlation matrix; S represents the trace of the genetic relationship matrix; n represents the reference population size; represents genetic variance; represents the residual variance; y represents the phenotypic value of the reference population; T represents transpose.

[0057] The formula for calculating the molecular marker effect value using the RHBLUP method is as follows:

[0058]

[0059] Here u represents the marker effect value; represents genetic variance; U represents the reference population marker matrix; T represents transposition; V -1 represents the inverse matrix of the sample variance of the reference population; y represents the phenotypic value of the reference population; A represents the fixed effect matrix;

[0060] (5) using the marker effect value obtained in step (4) to calculate the breeding value of each individual in the test population, and then sorting the breeding values ​​of each individual in the test population to select the best individual;

[0061] The breeding value calculation formula is as follows:

[0062] C=U c u

[0063] Here C represents the individual breeding value of the test population, U c represents the marker matrix of the test population, and u represents the marker effect value.

[0064] Test Example 1

[0065] The simulated population was calculated by the method of performing whole genome selection breeding based on representative populations by RHBLUP algorithm and GBLUP algorithm, respectively. The parameters of the population generated by computer simulation were as follows: double haploid (DH) population size was 13,000, the number of molecular markers was 17,001, the true heritability was 0.75, the test population size was 100, and the number of simulations was 10 times.

[0066] From the simulation results, it can be seen that the prediction accuracy of the method of the present invention for whole genome selection breeding based on representative populations through the RHBLUP algorithm is 0.70, and the calculation time for each simulation is 13 minutes; the prediction accuracy obtained by the GBLUP algorithm using all individuals is 0.73, and the calculation time for each simulation is 392 minutes. The results show that the calculation time of the method of the present invention for whole genome selection breeding based on representative populations through the RHBLUP algorithm is 30 times shorter than that of the GBLUP algorithm using all individuals, and the prediction accuracy is similar to that of the GBLUP algorithm using all individuals.

[0067] Test Example 2

[0068] The simulated population was calculated by the method of performing whole genome selection breeding based on representative populations by RHBLUP algorithm and GBLUP algorithm, respectively. The parameters of the population generated by computer simulation were as follows: the size of the hybrid population (F2) was 14,500, the number of molecular markers was 17,601, the true heritability was 0.73, the test population size was 100, and the number of simulations was 10 times.

[0069] From the simulation results, it can be seen that the prediction accuracy of the method of the present invention for whole genome selection breeding based on representative populations through the RHBLUP algorithm is 0.69, and the calculation time for each simulation is 15 minutes; the prediction accuracy obtained by the GBLUP algorithm using all individuals is 0.71, and the calculation time for each simulation is 482 minutes. The results show that the calculation time of the method of the present invention for whole genome selection breeding based on representative populations through the RHBLUP algorithm is 32 times shorter than that of the GBLUP algorithm using all individuals, and the prediction accuracy is similar to that of the GBLUP algorithm using all individuals.

[0070] The above two results show that the prediction accuracy of the method of the present invention for whole genome selection breeding based on representative populations through the RHBLUP algorithm is similar to that of the GBLUP method using all individuals, but the calculation efficiency is greatly improved.

[0071] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for whole genome selection breeding based on representative populations, characterized in that: The following steps are involved: S1, selecting a preset number of molecular markers in the reference population to construct an evolutionary tree, selecting multiple individuals at each node of the evolutionary tree, and merging the individuals extracted at all nodes to obtain population 1; S2, sort the individuals from high to low according to the phenotypic values ​​of the target traits of the reference population, then extract the individuals with the highest, middle and lowest phenotypic values ​​respectively, merge these individuals to obtain population 2; among them, the number of individuals with the highest, middle and lowest phenotypic values ​​is the same, accounting for 15-20% of the number of individuals in the reference population respectively; S3, merging population 1 and population 2 as the representative population, and using the whole genome selection algorithm RHBLUP algorithm to calculate the marker effect value of the representative population; S4, calculating the breeding value of each individual in the test population according to the marker effect value, and selecting the best individual according to the breeding value; Wherein, the step S3 uses the whole genome selection algorithm RHBLUP algorithm to calculate the representative population marker effect value, including: using the random HE method to estimate the genetic variance of the reference population; and using the BLUP method to calculate the molecular marker effect value based on the genetic variance of the reference population.

2. The method for whole genome selection breeding based on representative populations according to claim 1, characterized in that: In step S1, 300-600 molecular markers are randomly selected from the reference population to construct an evolutionary tree of the reference population, 3-5 individuals are randomly selected at each node of the evolutionary tree, and then merged to obtain population 1.

3. The method for whole genome selection breeding based on representative population according to claim 1, characterized in that: The formula for estimating the genetic variance of the reference population using the random HE method is: ; Where tr represents the trace of the matrix; represents the genetic correlation matrix; represents the trace of the genetic relationship matrix; represents the reference group size; represents genetic variance; represents the residual variance; represents the phenotypic value of the reference population; T represents transposition.

4. The method for whole genome selection breeding based on representative populations according to claim 1, characterized in that: The formula for calculating the molecular marker effect value by the BLUP method is: ;in, represents the marker effect value; represents genetic variance; represents the reference population marker matrix; T represents transpose; The inverse matrix representing the sample variance of the reference group; represents the phenotypic value of the reference population; represents the fixed effects matrix.

5. The method for whole genome selection breeding based on representative population according to claim 1, characterized in that: The calculation formula of the breeding value in step S4 is: ;in represents the individual breeding values ​​of the test population, represents the marker matrix of the test population, represents the marker effect value.

Images