Method for correcting performance score of lean type pigs with growth traits reaching 120kg

Abstract

The invention discloses a method for correcting performance scores of lean type pigs with the growth traits reaching 120kg, which comprises the following steps: collecting two-point data and measurement station data of lean type male and female pigs in a 90-140kg weight stage, performing preprocessing and quality control on the two-point data, then filtering the two-point data and the measurement station data, and correcting the performance scores of the lean type pigs with the growth traits reaching 120kg on the basis of the filtered two-point data and the measurement station data. Aiming at each growth character, fitting a growth model by using a nonlinear least square method, and selecting the growth model with the highest fitting degree mean value of each character on each variety and gender as a standard growth model; based on the fitted standard growth model of each character, fitting to obtain an individual growth curve; on the basis of the fitted individual growth curve, correction formula parameters are obtained through a unary linear regression method, and on the basis of the correction formula parameters, a 120 kg correction result is obtained through calculation according to a correction formula. The method can effectively improve correction phenotype accuracy and breeding value estimation accuracy.

Description

Technical Field

[0001] The invention belongs to the field of livestock and poultry growth trait correction, and in particular relates to a method for correcting the performance score of a lean pig with a growth trait of up to 120kg. Background Art

[0002] The correction of growth traits of breeding pigs, such as age, backfat thickness and eye muscle depth, can standardize the performance measurement data of breeding pigs, eliminate the weight differences between individuals, and make accurate performance comparisons of breeding pigs at the same weight level. The use of scientific and accurate correction methods to correct growth traits is a prerequisite for genetic evaluation of breeding pig growth traits and the basis for efficient breeding and genetic improvement of growth traits.

[0003] The National Swine Genetic Improvement Plan (2021-2035) sets the age of 120kg as the core breeding trait for breeding pigs. More and more breeding companies have begun to carry out large-weight performance testing and breeding of breeding pigs. However, there is currently no method for correcting growth traits up to 120kg in my country, making it difficult to carry out effective genetic evaluation and breeding for these three traits. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a method for correcting the performance results of growth traits of lean pigs up to 120kg. Based on the correction formula and parameter calculation method proposed in the present invention, the correction values ​​of the corrected age of pigs up to 120kg, the corrected back fat thickness up to 120kg and the corrected eye muscle depth up to 120kg can be accurately calculated, thereby realizing the individual comparison and genetic evaluation analysis of the growth performance of lean pigs up to 120kg in weight.

[0005] The present invention provides a method for correcting the performance of lean pigs with growth traits up to 120 kg, comprising:

[0006] Collecting two-point data and measuring station data of lean male and female pigs at a weight stage of 90 to 140 kg, preprocessing and quality controlling the two-point data, and then filtering the two-point data and measuring station data to obtain filtered two-point data and measuring station data;

[0007] Based on the filtered two-point data and the measuring station data, a growth model is fitted for each growth trait using a nonlinear least squares method, and the growth model with the highest mean fitting degree for each trait in each variety and gender is selected as the standard growth model;

[0008] Based on the standard growth model of each trait, the individual growth curve is fitted;

[0009] Based on the fitted individual growth curve, a univariate linear regression method was used to obtain correction formula parameters, and based on the correction formula parameters, a 120 kg correction result was obtained by calculation according to the correction formula.

[0010] Preferably, the process of preprocessing the two-point data includes:

[0011] (1) Delete the records of individuals with any missing pig individual number, gender, breed, or date of birth;

[0012] (2) Delete the records of all individuals involved in the duplication of pig individual numbers;

[0013] (3) Delete the records of individuals with only one measurement data.

[0014] Preferably, the process of quality control of the two-point data includes:

[0015] (1) The data should be logical: the weight, age, backfat thickness and eye muscle depth of the pig measured in the last measurement should be greater than the previous measurement;

[0016] (2) The data should be within the valid range: age, backfat thickness and eye loin depth should be quality controlled according to the following conditions;

[0017] For age traits, the first measured age of the pigs shall not be less than 100 days, and the last measured age shall not exceed 210 days. The first measured weight shall not be less than 80 kg, and the last measured weight shall not exceed 150 kg.

[0018] For the backfat thickness trait, the first measured backfat thickness shall not be less than 5mm, and the last measured backfat thickness shall not exceed 25mm; the first measured body weight shall not be less than 80kg, and the last measured body weight shall not exceed 150kg;

[0019] For the eye muscle depth trait, the eye muscle depth measured for the first time shall not be less than 30mm, and the eye muscle depth measured for the last time shall not exceed 85mm; the body weight measured for the first time shall not be less than 80kg, and the body weight measured for the last time shall not exceed 150kg;

[0020] (3) The growth rate of traits should be within the normal range: age, backfat thickness and eye loin depth should be determined according to the following conditions;

[0021] According to the two measurement records of the pigs, the average daily weight gain, back fat thickness increment per unit weight and eye muscle depth increment per unit weight of each individual were calculated by sex, and then the upper and lower quartiles and interquartile range of daily weight gain, back fat thickness increment per unit weight and eye muscle depth increment per unit weight were obtained. The upper and lower limits of normal data were defined according to the quartile method, and all individual data outside the normal data range were deleted;

[0022] (4) Body weight, age, backfat thickness and eye loin depth are all within a reasonable range: Age, backfat thickness and eye loin depth are quality controlled according to the following conditions;

[0023] The weight range was divided into 5-kg stages from 80 to 150 kg. The upper and lower quartiles and interquartile range of age, backfat thickness and eye muscle depth were calculated for each stage, and the upper and lower limits of normal data were defined according to the quartile method.

[0024] For age traits, the stages were divided into units of 10 days from 100 to 210 days;

[0025] For the backfat thickness trait, the stages are divided from 5 to 25 mm in units of 2 mm;

[0026] For eye muscle depth traits, the stages are divided into 5-mm units from 30 to 85 mm;

[0027] The upper and lower quartiles and interquartile range of each stage were calculated respectively, and the upper and lower limits of normal data were defined according to the quartile method, and the data of all individuals outside the normal data range were deleted.

[0028] Preferably, based on the filtered two-point data and the measuring station data, the growth model is fitted using the nonlinear least squares method for each growth trait, and the process of selecting the growth model with the highest mean fitting degree of each trait in each variety and gender as the standard growth model includes:

[0029] For age traits, fitting was performed based on all filtered measurement station data and two-point data; for backfat thickness and eye muscle depth traits, fitting was performed based on all filtered two-point data; the growth curves to be fitted included univariate linear, Saturation, Logistic, Gompertz and VonBertalanffy growth curves;

[0030] For age traits, age was the independent variable and body weight was the dependent variable;

[0031] For the traits of backfat thickness and eye loin depth, body weight was the independent variable, and backfat thickness and eye loin depth were the dependent variables;

[0032] Based on the above quality-controlled data, nonlinear least squares method was used for nonlinear regression modeling, and the growth model with the highest mean fitting degree of each trait in each breed and gender was selected as the standard growth model; among them, the standard growth model for age trait was the Gompertz model, the standard growth model for backfat thickness trait was the Saturation model, and the standard growth model for eye muscle depth trait was a univariate linear model.

[0033] Preferably, based on the fitted standard growth model of each trait, the process of fitting to obtain the individual growth curve comprises:

[0034] For the Saturation model of the standard growth model for backfat thickness and the univariate linear model of the standard growth model for eye muscle depth, the equations were solved directly based on two measured points to obtain the individual growth curves.

[0035] According to the Gompertz model of the standard growth model for day-age traits, the correction coefficient method was used to calculate the correction values ​​of each weight point from 90 to 140 kg, and the correction values ​​of each weight point from 90 to 140 kg were fitted to obtain the individual growth curve.

[0036] Preferably, the formula expression corrected to w0kg by the correction coefficient method is:

[0037]

[0038] w0kg correction value for a certain individual = trait value corresponding to the measured weight × w0kg correction coefficient

[0039] Preferably, based on the fitted individual growth curve, the process of obtaining correction formula parameters using a univariate linear regression method includes:

[0040] Based on the individual growth curve, a univariate linear regression equation is fitted, and the intercept of the regression equation is calculated, and the mean of all individual intercepts is the correction formula parameter;

[0041] Among them, in the process of fitting the univariate linear regression equation, for the age and eye muscle depth traits, body weight is the independent variable, and age and eye muscle depth are the dependent variables; for the back fat thickness trait, back fat thickness is the independent variable, and body weight is the dependent variable.

[0042] Preferably, the correction formula includes a 120kg age correction formula, a 120kg backfat thickness correction formula, and a 120kg eye muscle depth correction formula;

[0043] The expression of the 120kg age correction formula is:

[0044]

[0045] Among them, A is the correction formula parameter of age in days;

[0046] The expression of the 120kg backfat thickness correction formula is:

[0047]

[0048] Among them, B is the correction formula parameter of back fat thickness;

[0049] The expression of the 120kg eye muscle depth correction formula is:

[0050]

[0051] Among them, C is the correction formula parameter of eye muscle depth.

[0052] Preferably, the method further comprises verifying the accuracy of the correction phenotype of the correction formula parameters;

[0053] The process of verifying the accuracy of the correction formula parameters for the correction phenotype includes:

[0054] Divide the sample data set into a reference group and a validation group, and calculate the true value of each trait 120kg of the validation group;

[0055] Using the reference group and the correction formula parameter calculation method, a 120 kg correction formula for each trait is calculated based only on the reference group;

[0056] The 120kg correction formula obtained only based on the reference group, the Chinese 100kg correction formula and the Canadian formula are respectively used, and based on the measured data of each pig in the verification group, the correction value of each pig in the verification group to 120kg is calculated; wherein, the correction accuracy of the age and backfat thickness traits is compared and verified using the 120kg correction formula proposed by the present invention and the Chinese 100kg formula; the correction accuracy of the eye muscle depth trait is compared and verified using the 120kg correction formula proposed by the present invention and the Canadian formula;

[0057] The measured points of each trait were divided into stages of 5 kg from 90 to 140 kg, and the correlation evaluation index of each stage was calculated. The evaluation index of each weight stage was averaged to obtain the corrected phenotypic accuracy result; wherein the evaluation indexes included Pearson correlation, Spearman correlation, root mean square error and mean absolute error.

[0058] Preferably, based on the measured data of each pig in the validation group, the process of calculating the correction value of each pig in the validation group to 120 kg includes:

[0059] The 120kg correction formula based on the reference group only, the Chinese 100kg correction formula and the Canadian formula were used respectively, based on the pedigree data of corresponding breeds of pigs in several pig farms and the growth performance measurement data of various traits, to calculate the corrected age at 120kg, corrected backfat thickness at 120kg and corrected eye muscle depth at 120kg. The best linear unbiased estimation model was used to calculate the breeding value estimation accuracy of each trait by 5-fold cross validation method. Among them, the breeding value estimation accuracy of age and backfat thickness traits was compared and evaluated using the 120kg correction formula based on the reference group only and the Chinese 100kg formula; the breeding value estimation accuracy of eye muscle depth trait was compared and evaluated using the 120kg correction formula based on the reference group only and the Canadian formula.

[0060] Compared with the prior art, the present invention has the following advantages and technical effects:

[0061] The invention provides a method for correcting the performance of a lean meat pig's growth trait of up to 120 kg. The method comprises the following steps: collecting two-point data of male and female pigs of three lean meat breeds, namely Large White pigs, Landrace pigs and Duroc pigs, at a weight stage of 90 to 140 kg and measuring station data of the pigs throughout the whole process; and preprocessing and quality control of the two-point data, fitting a standard growth model, and calculating correction formula parameters through univariate linear regression. Correction formulas for corrected age, corrected backfat thickness and back muscle depth of the pigs reaching 120 kg are obtained. Correction using the correction formula can effectively improve the accuracy of the corrected phenotype and the accuracy of breeding value estimation, thereby solving the problem that there is no correction method for growth traits reaching 120 kg in my country and that effective genetic evaluation and breeding cannot be carried out for growth traits. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0063] Figure 1 A schematic diagram of a method flow of an embodiment of the present invention;

[0064] Figure 2 The frequency distribution histogram of daily weight gain of lean pigs according to an embodiment of the present invention;

[0065] Figure 3 The frequency distribution histogram of the increment of backfat thickness per unit body weight of lean pigs according to an embodiment of the present invention;

[0066] Figure 4 The frequency distribution histogram of the incremental depth of eye muscle per unit body weight of lean pigs according to an embodiment of the present invention;

[0067] Figure 5 This is a schematic diagram of a standard growth curve of lean pig age traits according to an embodiment of the present invention;

[0068] Figure 6 This is a schematic diagram of a standard growth curve of the backfat thickness trait of lean pigs according to an embodiment of the present invention;

[0069] Figure 7 Schematic diagram of the standard growth curve of the lean pig eye muscle depth trait according to an embodiment of the present invention. DETAILED DESCRIPTION

[0070] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0071] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0072] like Figure 1-7 As shown, this embodiment provides a method for correcting the performance of a lean pig with a growth trait of up to 120 kg, comprising:

[0073] Step 100, respectively collecting growth performance measurement data of any two weight points in the 90-140 kg weight stage of male and female pigs of three breeds, namely Large White pigs, Landrace pigs and Duroc pigs (hereinafter referred to as “two-point data”) and automatic feeding measurement station data of the pigs throughout the whole process (hereinafter referred to as “measurement station data”), wherein the measurement station data requires that the final weight of the pigs is greater than or equal to 100 kg;

[0074] The two data sources of this embodiment are the growth performance measurement data collected by the National Animal Husbandry Station from March 2023 to February 2024, involving 19,000 breeding pigs from 30 farms. The data source of the measurement station is the measurement station data submitted to the Animal Husbandry Comprehensive Information Platform by the National Core Breeding Farm from March 2023 to December 2023, involving 14,000 breeding pigs from 16 farms.

[0075] Step 200, preprocessing and quality control of the two-point data, including:

[0076] The data preprocessing process includes:

[0077] (1) Delete the records of individuals with any missing pig individual number, gender, breed, or date of birth;

[0078] (2) Delete the records of all individuals involved in the duplication of pig individual numbers;

[0079] (3) Delete the records of individuals with only one measurement data;

[0080] The quality control process includes:

[0081] (1) The data should be logical: the weight, age, backfat thickness and eye muscle depth of the pig measured in the last measurement should be greater than the previous measurement;

[0082] (2) The data should be within the valid range: The three traits were quality controlled separately according to the following conditions.

[0083] For age traits, the first measured age of the pigs shall not be less than 100 days, and the last measured age shall not exceed 210 days. The first measured weight shall not be less than 80 kg, and the last measured weight shall not exceed 150 kg.

[0084] For the backfat thickness trait, the first measured backfat thickness shall not be less than 5mm, and the last measured backfat thickness shall not exceed 25mm; the first measured body weight shall not be less than 80kg, and the last measured body weight shall not exceed 150kg;

[0085] For the eye muscle depth trait, the eye muscle depth measured for the first time shall not be less than 30mm, and the eye muscle depth measured for the last time shall not exceed 85mm; the body weight measured for the first time shall not be less than 80kg, and the body weight measured for the last time shall not exceed 150kg;

[0086] (3) The growth rate should be within the normal range: the three traits are quality controlled separately according to the following conditions.

[0087] According to the two measurement records of the pigs, the average daily weight gain, the increment of backfat thickness per unit weight and the increment of eye muscle depth per unit weight were calculated by sex, and then the upper and lower quartiles and interquartile ranges of these three indicators were obtained. The upper and lower limits of normal data were defined according to the quartile method (the upper limit of normal value is UQ+1.5IQR, the lower limit of normal value is LQ-1.5IQR, UQ, LQ and IQR are the upper quartile, lower quartile and middle quartile range respectively), and all individual data outside the normal data range were deleted; Figure 2 , Figure 3 and Figure 4 The frequency distribution histograms of daily weight gain, backfat thickness increment per unit weight and eye muscle depth increment per unit weight of breeding pigs are respectively illustrated; among them, Figure 2 A and B in the middle are sows and boars, respectively, and the dotted lines indicate the upper and lower limits of normal data; Figure 3 A and B are sows and boars, respectively, and the dotted line indicates the upper limit of normal data; Figure 4 A and B are sows and boars, respectively, and the dotted line indicates the upper limit of normal data;

[0088] (4) Body weight, age, back fat thickness and eye loin depth are all within a reasonable range: the three traits are quality controlled separately according to the following conditions.

[0089] The weight of the pigs was divided into stages with every 5 kg from 80 to 150 kg. The upper and lower quartiles and interquartile range of age, backfat thickness and eye muscle depth were calculated for each stage, and the upper and lower limits of normal data were defined according to the quartile method.

[0090] Similarly, for the age trait, the stages were divided into units of 10 days from 100 to 210 days;

[0091] For the backfat thickness trait, the stages are divided from 5 to 25 mm in units of 2 mm;

[0092] The eye muscle depth characteristics are divided into stages from 30 to 85 mm in units of 5 mm.

[0093] The upper and lower quartiles and interquartile range of each stage were calculated, and the upper and lower limits of normal data were defined according to the quartile method. All data of individuals outside the normal data range were deleted.

[0094] Step 300, based on the filtered two-point data and the measurement station data, fitting the standard growth model of each trait, specifically:

[0095] For the filtered two-point data and measuring station data, the nonlinear least squares method was used to fit multiple growth models for the three traits respectively, and the model with the highest mean fitting degree in each variety and gender was selected as the standard growth model for each trait.

[0096] Among them: For the age trait, fitting is performed based on all filtered measurement station data and two-point data,

[0097] For the backfat thickness and loin depth traits, fitting was performed based on all filtered two-point data.

[0098] The growth curves to be fitted include univariate linear, Saturation, Logistic, Gompertz and VonBertalanffy growth curves.

[0099] For age traits, age was the independent variable and body weight was the dependent variable;

[0100] For the traits of backfat thickness and loin depth, body weight was the independent variable, and backfat thickness and loin depth were the dependent variables.

[0101] Based on the above quality-controlled data, nonlinear regression modeling was performed using the nonlinear least squares method, and the model with the highest mean fitting degree in each variety and gender was selected as the standard growth model for each trait.

[0102] In this embodiment, the standard growth model for the age trait is the Gompertz model, the standard growth model for the backfat thickness trait is the Saturation model, and the standard growth model for the eye muscle depth trait is a univariate linear model. Figure 5 , Figure 6 and Figure 7 The standard growth models for age, backfat thickness, and eye loin depth are illustrated respectively. Figure 5 Middle AF are Large White sow, Large White boar, Landrace sow, Landrace boar, Duroc sow, and Duroc boar; Figure 6 Middle AF are Large White sow, Large White boar, Landrace sow, Landrace boar, Duroc sow, and Duroc boar; Figure 7 Middle AF are Large White sow, Large White boar, Landrace sow, Landrace boar, Duroc sow, and Duroc boar;

[0103] Step 400, based on the standard growth model of each trait, fitting to obtain an individual growth curve includes:

[0104] The standard growth model for backfat thickness is the Saturation model, and the standard growth model for eye muscle depth is a univariate linear model. The equation is solved directly based on two measured points to obtain the individual growth curve; the standard growth model for age is the Gompertz model, and the individual growth curve is calculated using the correction coefficient method. Specifically, the correction coefficient method can be expressed as follows, taking the correction to w0kg as an example:

[0105]

[0106] w0kg correction value for a certain individual = trait value corresponding to the measured weight × w0kg correction coefficient

[0107] Similarly, the correction values ​​for each weight point of 90 to 140 kg for the pig can be calculated to obtain the individual growth curve.

[0108] In this embodiment, the standard growth model for age traits is the Gompertz model, so the correction coefficient method is used to fit the individual growth curve. The standard growth models for backfat thickness and eye muscle depth traits are Saturation and univariate linear, respectively. Since there are only two parameters, the equation can be directly solved based on at least two measured points to obtain the individual growth curve.

[0109] Step 500, based on the individual growth curve, using a univariate linear regression method to calculate correction formula parameters, including:

[0110] Based on the individual growth curve, a univariate linear regression equation is fitted, and the intercept of the regression equation is calculated. The mean of the intercepts of all individuals is the correction formula parameter required. In the process of fitting a univariate linear regression equation, for the age and eye muscle depth traits, body weight is the independent variable, and age and eye muscle depth are the dependent variables; for the back fat thickness trait, back fat thickness is the independent variable, and body weight is the dependent variable. This embodiment calculates the correction formula parameters for the 120 kg body weight of male and female pigs of Large White Pigs, Landrace Pigs and Duroc Pigs. The correction formula parameters for age, back fat thickness and eye muscle depth are shown in Table 1, Table 2 and Table 3, respectively.

[0111] The correction formula for 120kg age is as follows:

[0112]

[0113] Among them, the A value is given in Table 1:

[0114] Table 1

[0115]

[0116] The correction formula for 120kg backfat thickness is as follows:

[0117]

[0118] Among them, the B value is given in Table 2:

[0119] Table 2

[0120]

[0121] The 120kg eye muscle depth correction formula is as follows:

[0122]

[0123] Among them, the C value is given in Table 3:

[0124] Table 3

[0125]

[0126] Furthermore, the correction formulas for the corrected age at 120 kg, the corrected backfat thickness at 120 kg, and the corrected eye muscle depth at 120 kg calculated according to this embodiment are compared with the correction formula of 100 kg in my country or the correction formula of the Canadian Pig Genetic Improvement Center (hereinafter referred to as the Canadian formula), and the correction phenotype accuracy of the correction formula parameters is verified, including:

[0127] (1) The sample data set is divided into a reference group and a validation group. Specifically, for age traits, the validation group is the data of at least 800 pigs with a measured weight ≥120 kg randomly selected from the measurement stations, and the reference group is the data of the pig measurement stations other than the validation group and all two-point data; for backfat thickness and eye muscle depth traits, in the two-point data, pigs with any measured record between [105,135] kg and the first measured weight <120 kg and the last measured weight >120 kg are selected, and individuals accounting for 20% of the total data are randomly selected as the validation group, and the remaining 80% of the two-point data are used as the reference group.

[0128] (2) Calculate the true value of 120 kg for each trait of the validation group. Specifically, for the age trait, use linear interpolation to obtain the true value of 120 kg body weight based on the two data points at the left and right ends of 120 kg and closest to 120 kg in the measurement station data; for the back fat thickness and eye muscle depth traits, calculate the true value of 120 kg body weight for each trait of each pig in the validation group based on the individual growth curve obtained.

[0129] (3) Using the reference group and the correction formula parameter calculation method, a 120 kg correction formula for each trait is calculated based only on the reference group.

[0130] (4) using the 120kg correction formula obtained only based on the reference group, the Chinese 100kg correction formula and the Canadian formula respectively, and based on the measured data of each pig in the verification group, calculating the correction value of each pig in the verification group to 120kg; wherein, the correction accuracy of the age and backfat thickness traits is compared and verified using the 120kg correction formula proposed by the present invention and the Chinese 100kg formula; the correction accuracy of the eye muscle depth trait is compared and verified using the 120kg correction formula proposed by the present invention and the Canadian formula.

[0131] (5) The calculation method of corrected phenotypic accuracy is as follows: the measured points of each trait are divided into stages of 5 kg from 90 to 140 kg, and the correlation evaluation index of each stage is calculated. The evaluation indexes include Pearson correlation, Spearman correlation, root mean square error (RMSE) and mean absolute error (MAE). The average of the evaluation indexes of each weight stage is the corrected phenotypic accuracy result.

[0132] In this embodiment, based on the correction formula parameters of the male and female Large White pigs, Landrace pigs and Duroc pigs reaching 120 kg body weight given in Tables 1, 2 and 3, the corrected phenotypic accuracy of each trait was calculated. The corrected phenotypic accuracy of age, backfat thickness and eye muscle depth are shown in Tables 4, 5 and 6, respectively.

[0133] Table 4

[0134]

[0135] Table 5

[0136]

[0137] Table 6

[0138]

[0139] Furthermore, according to the correction formulas of the corrected age at 120 kg, the corrected backfat thickness at 120 kg and the corrected eye muscle depth at 120 kg calculated in this embodiment, the correction formula parameters are compared with the 100 kg correction formula of my country or the Canadian formula to evaluate the accuracy of breeding value estimation, including:

[0140] The 120kg correction formula obtained by the correction method provided by the present invention, the 100kg correction formula of my country and the Canadian formula are respectively used, based on the pedigree data of the corresponding breed pigs of several pig farms and the growth performance measurement data of each trait, to calculate the corrected age of up to 120kg, the corrected backfat thickness of up to 120kg and the corrected eye muscle depth of up to 120kg, and the best linear unbiased estimation model is used to calculate the breeding value estimation accuracy of each trait by a 5-fold cross validation method. The breeding value estimation accuracy of the age and backfat thickness traits is compared and evaluated using the 120kg correction formula proposed by the present invention and the 100kg formula of my country; the breeding value estimation accuracy of the eye muscle depth trait is compared and evaluated using the 120kg correction formula proposed by the present invention and the Canadian formula.

[0141] Specifically, the age and backfat thickness adopt a double-trait animal model, and the genetic evaluation model is shown as follows:

[0142]

[0143] Among them, y1 and y2 are the vectors of corrected age at 120 kg and corrected back fat thickness at 120 kg; b1 and b2 are the fixed effect vectors of the two traits, namely, the fixed effects of field year and season and sex; l1 and l2 are the random litter effect vectors of the two traits; a1 and a2 are the additive genetic effect vectors of the two traits; e1 and e2 are the random environmental effect vectors of the two traits; X1, X2, W1, W2, Z1, and Z2 are the corresponding association matrices.

[0144] The eye muscle depth adopts a single trait animal model, and the genetic evaluation model is as follows:

[0145] y=Xb+Wl+Za+e

[0146] Among them, y is the vector of corrected eye muscle depth up to 120 kg; b is the fixed effect vector, that is, the field year season and sex fixed effects; l is the random litter effect vector; a is the additive genetic effect vector; e is the random environmental effect vector; X, W, Z are the corresponding association matrices.

[0147] Specifically, the method for calculating the accuracy of breeding value estimation by 5-fold cross validation is as follows: First, based on the entire data set, the corrected phenotypic value y of the individual with no missing phenotypic data in the pedigree is calculated. c . Then, 20% of the phenotypic data of the individuals were randomly removed and the estimated breeding values ​​EBV of these individuals were calculated. The accuracy of the breeding value estimation was calculated as the correlation coefficient between the estimated breeding value and the corrected phenotypic value. The above process was repeated using a five-fold cross-validation method.

[0148]

[0149] This embodiment calculates the estimated accuracy of breeding values ​​for each trait based on the correction formula parameters for male and female Large White, Landrace and Duroc pigs reaching a body weight of 120 kg given in Tables 1, 2 and 3. The corrected phenotypic accuracy of age, backfat thickness and eye muscle depth is shown in Table 7.

[0150] Table 7

[0151]

[0152] In this embodiment, by collecting two-point data of 90-140kg weight stage of male and female pigs of three breeds of Large White Pig, Landrace Pig and Duroc respectively and the whole process of pig measurement station data, and pre-processing and quality control of the two-point data, the standard growth model is fitted using nonlinear least squares method; based on the standard growth model of each trait, the individual growth curve is fitted; based on the individual growth curve, the correction formula parameters are calculated using univariate linear regression method. Using the performance score correction method of lean meat pig growth trait up to 120kg provided by the present invention, more accurate correction formulas for the three traits of corrected age, corrected backfat thickness and corrected eye muscle depth of lean meat pig breeds up to 120kg can be obtained, which closely fits the trend of large-weight selection and breeding of lean meat pigs in my country, improves the pig genetic improvement technology system with independent characteristics in my country, enriches the breeding target traits of lean meat pig breeds in my country, and plays an important technical support role in accelerating the second round of pig genetic improvement plan.

[0153] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for correcting the performance of lean pigs growing to 120 kg, characterized in that: include: Collecting two-point data and measuring station data of lean male and female pigs at a weight stage of 90 to 140 kg, preprocessing and quality controlling the two-point data, and then filtering the two-point data and measuring station data to obtain filtered two-point data and measuring station data; Based on the filtered two-point data and the measuring station data, a growth model is fitted for each growth trait using a nonlinear least squares method, and the growth model with the highest mean fitting degree for each trait in each variety and gender is selected as the standard growth model; Based on the standard growth model of each trait, the individual growth curve is fitted; Based on the fitted individual growth curve, a univariate linear regression method was used to obtain correction formula parameters, and based on the correction formula parameters, a 120 kg correction result was obtained by calculation according to the correction formula.

2. The method for correcting the performance of lean pigs with growth characteristics up to 120 kg according to claim 1, characterized in that: The process of preprocessing the two-point data includes: (1) Delete the records of individuals with any missing pig individual number, gender, breed, or date of birth; (2) Delete the records of all individuals involved in the duplication of pig individual numbers; (3) Delete the records of individuals with only one measurement data.

3. The method for correcting the performance of lean pigs with growth characteristics up to 120 kg according to claim 1, characterized in that: The process of quality control of the two-point data includes: (1) The data should be logical: the weight, age, backfat thickness and eye muscle depth of the pig measured in the last measurement should be greater than the previous measurement; (2) The data should be within the valid range: age, backfat thickness and eye loin depth should be quality controlled according to the following conditions; For age traits, the first measured age of the pigs shall not be less than 100 days, and the last measured age shall not exceed 210 days. The first measured weight shall not be less than 80 kg, and the last measured weight shall not exceed 150 kg. For the backfat thickness trait, the first measured backfat thickness shall not be less than 5mm, and the last measured backfat thickness shall not exceed 25mm; the first measured body weight shall not be less than 80kg, and the last measured body weight shall not exceed 150kg; For the eye muscle depth trait, the eye muscle depth measured for the first time shall not be less than 30mm, and the eye muscle depth measured for the last time shall not exceed 85mm; the body weight measured for the first time shall not be less than 80kg, and the body weight measured for the last time shall not exceed 150kg; (3) The growth rate of traits should be within the normal range: age, backfat thickness and eye loin depth should be determined according to the following conditions; According to the two measurement records of the pigs, the average daily weight gain, back fat thickness increment per unit weight and eye muscle depth increment per unit weight of each individual were calculated by sex, and then the upper and lower quartiles and interquartile range of daily weight gain, back fat thickness increment per unit weight and eye muscle depth increment per unit weight were obtained. The upper and lower limits of normal data were defined according to the quartile method, and all individual data outside the normal data range were deleted; (4) Body weight, age, backfat thickness and eye loin depth are all within a reasonable range: Age, backfat thickness and eye loin depth are quality controlled according to the following conditions; The weight range was divided into 5-kg stages from 80 to 150 kg. The upper and lower quartiles and interquartile range of age, backfat thickness and eye muscle depth were calculated for each stage, and the upper and lower limits of normal data were defined according to the quartile method. For age traits, the stages were divided into units of 10 days from 100 to 210 days; For the backfat thickness trait, the stages are divided from 5 to 25 mm in units of 2 mm; For eye muscle depth traits, the stages are divided into 5-mm units from 30 to 85 mm; The upper and lower quartiles and interquartile range of each stage were calculated respectively, and the upper and lower limits of normal data were defined according to the quartile method, and the data of all individuals outside the normal data range were deleted.

4. The method for correcting the performance of lean pigs growing to 120 kg according to claim 1, characterized in that: Based on the filtered two-point data and the measuring station data, the process of fitting the growth model for each growth trait using the nonlinear least squares method and selecting the growth model with the highest mean fitting degree for each trait in each variety and gender as the standard growth model includes: For age traits, fitting was performed based on all filtered measurement station data and two-point data; for backfat thickness and eye muscle depth traits, fitting was performed based on all filtered two-point data; the growth curves to be fitted included univariate linear, Saturation, Logistic, Gompertz and VonBertalanffy growth curves; For age traits, age was the independent variable and body weight was the dependent variable; For the traits of backfat thickness and eye loin depth, body weight was the independent variable, and backfat thickness and eye loin depth were the dependent variables; Based on the above quality-controlled data, nonlinear least squares method was used for nonlinear regression modeling, and the growth model with the highest mean fitting degree of each trait in each breed and gender was selected as the standard growth model; among them, the standard growth model for age trait was the Gompertz model, the standard growth model for backfat thickness trait was the Saturation model, and the standard growth model for eye muscle depth trait was a univariate linear model.

5. The method for correcting the performance of lean pigs with growth characteristics up to 120 kg according to claim 1, characterized in that: Based on the fitted standard growth model of each trait, the process of fitting the individual growth curve includes: For the Saturation model of the standard growth model for backfat thickness and the univariate linear model of the standard growth model for eye muscle depth, the equations were solved directly based on two measured points to obtain the individual growth curves. According to the Gompertz model of the standard growth model for day-age traits, the correction coefficient method was used to calculate the correction values ​​of each weight point from 90 to 140 kg, and the correction values ​​of each weight point from 90 to 140 kg were fitted to obtain the individual growth curve.

6. The method for correcting the performance of lean pigs growing to 120 kg according to claim 5, characterized in that: The formula expression corrected to w0kg by the correction coefficient method is: The w0kg correction value of an individual = the trait value corresponding to the actual body weight × w0kg correction coefficient.

7. The method for correcting the performance of lean pigs growing to 120 kg according to claim 1, characterized in that: Based on the fitted individual growth curve, the process of obtaining the correction formula parameters using the univariate linear regression method includes: Based on the individual growth curve, a univariate linear regression equation is fitted, and the intercept of the regression equation is calculated, and the mean of all individual intercepts is the correction formula parameter; Among them, in the process of fitting the univariate linear regression equation, for the age and eye muscle depth traits, body weight is the independent variable, and age and eye muscle depth are the dependent variables; for the back fat thickness trait, back fat thickness is the independent variable, and body weight is the dependent variable.

8. The method for correcting the performance of lean pigs with growth characteristics up to 120 kg according to claim 1, characterized in that: The correction formulas include a 120kg age correction formula, a 120kg backfat thickness correction formula, and a 120kg eye muscle depth correction formula; The expression of the 120kg age correction formula is: Among them, A is the correction formula parameter of age in days; The expression of the 120kg backfat thickness correction formula is: Among them, B is the correction formula parameter of back fat thickness; The expression of the 120kg eye muscle depth correction formula is: Among them, C is the correction formula parameter of eye muscle depth.

9. The method for correcting the performance of lean pigs with growth traits up to 120 kg according to claim 1, characterized in that: The method further comprises verifying the accuracy of the correction phenotype of the correction formula parameters; The process of verifying the accuracy of the correction formula parameters for the correction phenotype includes: Divide the sample data set into a reference group and a validation group, and calculate the true value of each trait 120kg of the validation group; Using the reference group and the correction formula parameter calculation method, a 120 kg correction formula for each trait is calculated based only on the reference group; The 120kg correction formula obtained only based on the reference group, the Chinese 100kg correction formula and the Canadian formula are respectively used, and based on the measured data of each pig in the verification group, the correction value of each pig in the verification group to 120kg is calculated; wherein, the correction accuracy of the age and backfat thickness traits is compared and verified using the 120kg correction formula proposed by the present invention and the Chinese 100kg formula; the correction accuracy of the eye muscle depth trait is compared and verified using the 120kg correction formula proposed by the present invention and the Canadian formula; The measured points of each trait were divided into stages of 5 kg from 90 to 140 kg, and the correlation evaluation index of each stage was calculated. The evaluation index of each weight stage was averaged to obtain the corrected phenotypic accuracy result; wherein the evaluation indexes included Pearson correlation, Spearman correlation, root mean square error and mean absolute error.

10. The method for correcting the performance of lean pigs growing to 120 kg according to claim 9, characterized in that: Based on the measured data of each pig in the validation group, the process of calculating the correction value for each pig in the validation group to 120 kg includes: The 120kg correction formula based on the reference group only, the Chinese 100kg correction formula and the Canadian formula were used respectively, based on the pedigree data of corresponding breeds of pigs in several pig farms and the growth performance measurement data of various traits, to calculate the corrected age at 120kg, corrected backfat thickness at 120kg and corrected eye muscle depth at 120kg. The best linear unbiased estimation model was used to calculate the breeding value estimation accuracy of each trait using the 5-fold cross-validation method. Among them, the breeding value estimation accuracy of age and backfat thickness traits was compared and evaluated using the 120kg correction formula based on the reference group only and the Chinese 100kg formula; the breeding value estimation accuracy of eye muscle depth trait was compared and evaluated using the 120kg correction formula based on the reference group only and the Canadian formula.

Images