A method for evaluating and analyzing genomic breeding values ​​of multiple beef cattle breeds

Abstract

The present invention relates to a method for evaluating and analyzing the genomic breeding values ​​of multiple varieties of beef cattle, and relates to the field of beef cattle breeding, including: generating multiple meat pattern maps based on genotype data and meat pattern map generation accuracy; configuring growth rate prediction accuracy, and performing growth rate phenotype prediction based on genotype data to obtain growth rate; matching and identifying multiple meat pattern maps and growth rates to obtain multiple matching degrees, and calculating and obtaining evaluation accuracy based on the meat pattern map generation accuracy and growth rate prediction accuracy; obtaining breeding values ​​based on multiple meat pattern maps and growth rate evaluations, and compensating the breeding values ​​using the evaluation accuracy to obtain breeding value intervals. This application can solve the technical problems of the traditional beef cattle breeding value evaluation method based on genetic data analysis, which lacks an accurate scientific basis and has insufficient evaluation accuracy and reliability; it can more comprehensively and scientifically evaluate the breeding potential of beef cattle, and significantly improve the scientificity, accuracy and reliability of beef cattle breeding value evaluation.

Description

Technical Field

[0001] The present invention relates to the field of beef cattle breeding, and in particular to a method for evaluating and analyzing genomic breeding values ​​of multiple varieties of beef cattle. Background Art

[0002] Beef cattle breeding is an important part of agricultural production to improve the production performance, meat quality and economic benefits of beef cattle. The purpose of breeding is to improve the growth rate, meat quality, reproductive ability and other economic traits of the next generation of beef cattle by selecting beef cattle with excellent genes for breeding, thereby improving breeding efficiency.

[0003] At present, the traditional beef cattle breeding value assessment method based on genetic data analysis mainly uses genetic characteristics as the assessment basis; however, the breeding value of beef cattle is not only affected by the genotype, but also by phenotypic characteristics such as meat quality, fat distribution, and muscle development, which have an important impact on the overall quality and production performance of beef cattle. Traditional methods fail to fully consider these factors, resulting in insufficient comprehensiveness and accuracy of the assessment results. Summary of the Invention

[0004] The present invention aims to solve the technical problems that traditional beef cattle breeding value evaluation methods based on genetic data analysis lack precise scientific basis and have insufficient evaluation accuracy and reliability. It provides a beef cattle multi-breed genomic breeding value evaluation and analysis method to solve the problem.

[0005] The present invention provides a method for evaluating and analyzing genomic breeding values ​​of multiple beef cattle breeds, comprising: using a gene chip to obtain genotype data of a target beef cattle and obtaining the beef cattle breed of the target beef cattle; configuring a meat texture pattern generation accuracy according to the beef cattle breed, and generating multiple meat texture pattern patterns according to the genotype data and the meat texture pattern generation accuracy; configuring a growth rate prediction accuracy according to the beef cattle breed, performing a growth rate phenotype prediction according to the genotype data, and predicting the growth rate; matching and identifying the multiple meat texture pattern patterns and growth rates to obtain multiple matching degrees, and calculating an evaluation accuracy based on the meat texture pattern generation accuracy and the growth rate prediction accuracy; evaluating and calculating a breeding value based on the multiple meat texture pattern patterns and growth rates, compensating the breeding value using the evaluation accuracy, and obtaining a breeding value interval as a breeding value analysis result.

[0006] Optionally, the method for evaluating and analyzing the genomic breeding values ​​of multiple beef cattle breeds further includes: using a gene chip to sequence the target beef cattle to obtain genotype data; and obtaining the beef cattle breed of the target beef cattle.

[0007] Optionally, the method for evaluating and analyzing the genomic breeding values ​​of multiple varieties of beef cattle also includes: obtaining the total breeding number of beef cattle and the breeding number of the beef cattle breeds; calculating the ratio of the breeding number to the total breeding number to obtain a breeding number coefficient; training an integrated number Q of meat pattern generation paths to obtain a meat pattern generation channel, where Q is a positive integer; calculating a meat pattern generation accuracy P based on the breeding number coefficient and Q, where P is a positive integer less than or equal to Q; and inputting the genotype data into P randomly selected meat pattern generation paths to generate P meat pattern maps.

[0008] Optionally, the method for evaluating and analyzing the genomic breeding values ​​of multiple varieties of beef cattle also includes: collecting a sample genotype data set and a sample meat quality pattern map set based on beef meat quality detection data in a historical period; randomly collecting Q copies of the sample genotype data set and the sample meat quality pattern map set with replacement to generate supervised training data; constructing Q meat quality pattern map generation paths based on a generative adversarial network, each meat quality pattern map generation path including a generator and a discriminator; using the Q copies to generate supervised training data respectively, and performing alternating iterative training on the generators and discriminators in the Q meat quality pattern map generation paths until the training is completed, and combining to obtain a meat quality pattern map generation channel.

[0009] Optionally, the method for evaluating and analyzing the genomic breeding values ​​of multiple varieties of beef cattle also includes: obtaining the total breeding number of beef cattle and the breeding number of the beef cattle breeds, and calculating the breeding number coefficient; training an integrated number Q of growth rate prediction paths to obtain a growth rate prediction channel, wherein the input data in the training process of each growth rate prediction path is genotype data, and the supervision data is the growth rate; based on the breeding number coefficient and Q, calculating the growth rate prediction accuracy P, where P is a positive integer less than or equal to Q; inputting the genotype data into P randomly selected growth rate prediction paths, predicting and outputting P predicted growth rates, and calculating the mean to obtain the growth rate.

[0010] Optionally, the method for evaluating and analyzing the genomic breeding values ​​of multiple varieties of beef cattle also includes: calculating a plurality of fat uniformity parameters based on the fat distribution in the plurality of meat texture patterns; inputting the plurality of fat uniformity parameters into a meat texture growth rate mapper to map and obtain a plurality of mapped growth rates, wherein the meat texture growth rate mapper includes a mapping relationship between sample fat uniformity parameters and sample growth rates; calculating a plurality of matching degrees based on the deviations between the plurality of mapped growth rates and the growth rates, and calculating the matching accuracy; calculating the basic accuracy based on the path accuracy of the meat texture pattern generation path and the growth rate prediction path selected according to the meat texture pattern generation accuracy and the growth rate prediction accuracy; and calculating the evaluation accuracy based on the matching accuracy and the basic accuracy.

[0011] Optionally, the method for evaluating and analyzing the genomic breeding values ​​of multiple breeds of beef cattle also includes: calculating the ratio of the growth rate and multiple mapped growth rates, and calculating the mean to obtain the growth rate breeding value; calculating the ratio of the multiple fat uniformity parameters and the preset fat uniformity parameters to obtain the fat breeding value; and weightedly calculating the fat breeding value and the growth rate breeding value to obtain the breeding value.

[0012] Optionally, the method for evaluating and analyzing genomic breeding values ​​of multiple beef cattle breeds further includes: calculating a compensation coefficient based on the evaluation accuracy; and performing interval compensation on the breeding value using the compensation coefficient to obtain a breeding value interval as a breeding value analysis result.

[0013] The beneficial effects of the present invention are as follows: by sequencing target beef cattle using a gene chip, the genotype data of the target beef cattle are obtained, and the beef cattle breed of the target beef cattle is obtained; then, according to the beef cattle breed, the meat pattern generation accuracy is configured, and according to the genotype data and the meat pattern generation accuracy, a plurality of meat pattern maps are generated; on the other hand, according to the beef cattle breed, the growth rate prediction accuracy is configured, and the growth rate phenotype is predicted according to the genotype data to predict the growth rate; further, the plurality of meat pattern maps and the growth rate are matched and identified to obtain a plurality of The matching degree is combined with the meat pattern generation accuracy and growth rate prediction accuracy to calculate the evaluation accuracy; then, based on the multiple meat pattern patterns and growth rates, the breeding value is evaluated and calculated; finally, the breeding value is compensated using the evaluation accuracy to obtain the breeding value interval as the breeding value analysis result; that is, by conducting multi-dimensional analysis based on the meat pattern pattern and the growth rate of beef cattle, and combining machine learning and other methods for intelligent evaluation, the breeding potential of beef cattle can be evaluated more comprehensively and scientifically, and the scientificity, accuracy and reliability of the breeding value evaluation of beef cattle can be significantly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 A schematic diagram of a process for evaluating and analyzing genomic breeding values ​​of multiple beef cattle breeds provided by the present invention;

[0015] Figure 2 A schematic diagram of a process for obtaining multiple meat quality pattern maps in a method for evaluating and analyzing genomic breeding values ​​of multiple beef cattle breeds provided by the present invention. DETAILED DESCRIPTION

[0016] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0017] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the specified features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0018] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or illustration". Any embodiment of the present invention described as "for example" is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed herein.

[0019] Examples, such as Figure 1 As shown, the embodiment of the present invention provides a method for evaluating and analyzing the genomic breeding values ​​of multiple breeds of beef cattle, which specifically includes the following steps:

[0020] S100: Using a gene chip, obtaining genotype data of a target beef cattle and obtaining the beef cattle breed of the target beef cattle.

[0021] Furthermore, step S100 of the present invention further includes:

[0022] S110: using a gene chip to sequence the target beef cattle to obtain genotype data; S120: obtaining the beef cattle breed of the target beef cattle.

[0023] Specifically, using gene chips to sequence target beef cattle and obtain genotype data is an effective means in modern beef cattle breeding research. Gene chip technology can quickly obtain genomic information of target beef cattle through high-throughput sequencing, providing a large amount of gene locus data, and providing a scientific basis for accurately evaluating the genetic potential and breeding value of beef cattle.

[0024] First, the genome of the target beef cattle is sequenced using a gene chip. By measuring specific gene markers in the beef cattle genome, the gene chip can accurately capture genetic information related to the production performance of beef cattle (such as growth rate, meat quality, reproductive ability, etc.), and obtain the genotype data of the target beef cattle. The genotype data includes single nucleotide polymorphisms, genomic marker sites, and information related to the association between genotype and phenotype. Among them, the gene chip can detect thousands of gene sites simultaneously, and compared with traditional genetic testing methods, it can provide more comprehensive genomic data. On the other hand, the breed of the target beef cattle can be obtained by comparing the genotype of the target beef cattle with the known breed genomic data using an existing genetic database to accurately determine the breed category of the beef cattle. Among them, different breeds of beef cattle may show obvious differences in meat quality, growth rate, fat distribution, etc.

[0025] S200: configuring a meat texture pattern generation accuracy according to the beef cattle breed, and generating a plurality of meat texture pattern maps according to the genotype data and the meat texture pattern generation accuracy.

[0026] Further, if Figure 2 As shown, step S200 of the present invention further includes:

[0027] S210: Obtain the total breeding number of beef cattle and the breeding number of the beef cattle breed; S220: Calculate the ratio of the breeding number to the total breeding number to obtain a breeding number coefficient.

[0028] Specifically, the total number of beef cattle raised is obtained, which refers to the total number of all beef cattle in a specific area (such as a farm); and the number of beef cattle breeds raised is obtained, which refers to the number of beef cattle of a specific breed raised in a specific area. Next, the ratio of the breeding number to the total breeding number is set as the breeding number coefficient. The breeding number coefficient reflects the proportion of the beef cattle breed in the entire breeding population. A high breeding number coefficient indicates that the breed accounts for a large proportion of the total breeding population, which means that the breeding of the breed in the area is relatively common or has high economic benefits; a low breeding number coefficient indicates that the breed is relatively unpopular.

[0029] S230: Training an integrated number Q of meat pattern generation paths to obtain a meat pattern generation channel, where Q is a positive integer.

[0030] Furthermore, step S230 of the present invention further includes:

[0031] S231: Based on the beef quality detection data in the historical period, a sample genotype data set and a sample meat pattern map set are collected; S232: Q copies are randomly collected with replacement from the sample genotype data set and the sample meat pattern map set to generate supervised training data; S233: Based on the generative adversarial network, Q meat pattern map generation paths are constructed, and each meat pattern map generation path includes a generator and a discriminator; S234: The Q copies are used to generate supervised training data respectively, and the generators and discriminators in the Q meat pattern map generation paths are alternately iteratively trained until the training is completed, and a meat pattern map generation channel is obtained by combination.

[0032] Specifically, first, the beef quality test data of the beef cattle breed in the historical time (such as the last year) are obtained. These data usually include measurement indicators of meat texture, fat distribution, muscle development, etc.; then, based on the beef quality test data in the historical time, a sample genotype data set and a sample meat pattern map set are collected, wherein the sample genotype data and the sample meat pattern map set are Figure 1 Correspondingly, the meat pattern diagram can intuitively show the texture structure and fat distribution of beef. Marble pattern is an important sign of high-quality beef. Usually, fat is evenly distributed in the meat layer, forming white patterns similar to snowflakes. The number, density and uniformity of distribution of these patterns largely determine the quality of the meat.

[0033] Then, the sample genotype data set and the sample meat pattern map set are used as sample data sets, and the sample data sets are divided into Q equal parts to obtain Q data sets, where Q is a positive integer and the specific value of Q can be set according to actual needs, such as being set to 20; further, Q data sets are selected with replacement Q times to obtain the first supervised training data, and the same method is used to iteratively select Q times to obtain Q generated supervised training data.

[0034] Then, based on the generative adversarial network, Q meat pattern generation paths are constructed. Each meat pattern generation path includes a generator and a discriminator. The generator is responsible for generating a network that simulates real data (such as meat pattern images), and the discriminator is responsible for judging whether the generated data is consistent with the real data, that is, whether it is "fake" data. In the application of meat pattern generation, the generator combines genotype data with meat pattern features and generates meat pattern images through deep learning technology, while the discriminator compares the real snowflake pattern image with the generated image to judge its authenticity.

[0035] Furthermore, the Q portions of supervised training data are used to perform alternating iterative training on the generator and discriminator within the Q meat texture pattern generation paths. The generator receives genotype data and corresponding meat texture pattern features as input and uses this information to generate simulated meat texture patterns. The discriminator's input is a meat texture pattern generated by the generator and a real meat texture pattern (sample meat texture pattern). The discriminator's goal is to determine whether each image is real data. That is, it determines the difference between the generated meat texture pattern and the real meat texture pattern, and outputs a true or false judgment result. During each training session, the generator and discriminator are updated alternately. First, the discriminator is trained using the real meat texture pattern and the generated image, enabling it to distinguish between real and generated images. The trained discriminator is then used to train the generator, allowing the generator to generate more realistic meat texture patterns by adjusting weights. During the training process, the loss functions of the generator and discriminator influence each other and are optimized alternately. The generator gradually optimizes the generated meat pattern by trying to make the discriminator output incorrect results. The discriminator, on the other hand, forces the generator to improve by increasing its judgment accuracy. By the end of training, the generator is able to generate high-quality meat pattern images based on genotype data, and the discriminator can effectively distinguish generated images from real images. The Q trained meat pattern generation paths are obtained. Finally, the meat pattern generation channel is obtained by combining the Q trained meat pattern generation paths.

[0036] By using a training method based on generative adversarial networks and combining it with Q-portion supervised training data, a highly accurate meat pattern generation model can be established. This model can generate high-quality meat pattern maps based on the genotype of beef cattle, thereby providing effective data support for the entire breeding analysis process and improving the scientificity and accuracy of breeding value assessment.

[0037] S240: Calculate the meat pattern generation accuracy P based on the breeding quantity coefficient and Q, where P is a positive integer less than or equal to Q; S250: Input the genotype data into P randomly selected meat pattern generation paths to generate P meat pattern maps.

[0038] Specifically, the breeding population coefficient is multiplied by Q and rounded to obtain a meat pattern generation accuracy P, where P is a positive integer less than or equal to Q. For example, assuming the breeding population coefficient is 18% and Q is 20, P is 0.18 multiplied by 20, rounded to 4. Next, P meat pattern generation paths are randomly selected from the Q meat pattern generation paths, and the genotype data is input into the P randomly selected meat pattern generation paths to output P meat pattern maps.

[0039] By setting the meat pattern generation accuracy according to the breeding quantity coefficient of the beef cattle breed, and selecting an appropriate number of meat pattern generation paths for analysis based on the meat pattern generation accuracy, that is, if the breeding quantity of a certain breed is large, its generation paths can be set more, thereby improving the diversity and quality of the generated image; on the contrary, if the breeding quantity of a certain breed is small, the number of generation paths can be appropriately reduced to avoid over-calculation; thereby, while ensuring the analysis accuracy, computing power resources can be saved and analysis efficiency can be improved.

[0040] S300: configuring growth rate prediction accuracy according to the beef cattle breed, performing growth rate phenotype prediction based on the genotype data, and predicting the growth rate.

[0041] Furthermore, step S300 of the present invention further includes:

[0042] S310: Obtain the total breeding quantity of beef cattle and the breeding quantity of the beef cattle breed, and calculate the breeding quantity coefficient; S320: Train the growth rate prediction path of the integrated quantity Q to obtain the growth rate prediction channel, wherein the input data in the training process of each growth rate prediction path is genotype data, and the supervision data is the growth rate; S330: Based on the breeding quantity coefficient and Q, calculate the growth rate prediction accuracy P, where P is a positive integer less than or equal to Q; S340: Input the genotype data into P randomly selected growth rate prediction paths, obtain P predicted growth rates through prediction output, and calculate the mean to obtain the growth rate.

[0043] Specifically, the total number of beef cattle raised is obtained, as is the number of the breed of beef cattle raised. The ratio of the number of the breed of beef cattle raised to the total number of cattle raised is then set as the breeding coefficient. Furthermore, based on historical beef cattle growth data, a set of sample genotype data and a set of sample growth rates are collected, where the sample genotype data and the sample growth rate correspond one-to-one. Q growth rate prediction paths are then constructed based on a generative adversarial network. Each growth rate prediction path includes a generator and a discriminator. The generator receives genotype data as input and outputs a predicted growth rate. The generator learns the underlying relationship between genotype and growth rate, generating growth rate prediction results that are as realistic as possible. The discriminator is responsible for determining whether the growth rate prediction output by the generator is realistic. It compares the generated growth rate with the actual growth rate data and provides feedback to the generator to optimize the generation process.

[0044] Then, the sample genotype data set and the sample growth rate set are used as sample data sets and divided into Q equal parts. The data sets are selected Q times with replacement to obtain the first training set, and the data sets are selected iteratively Q times to obtain Q training sets. The Q training sets are then used to perform alternating iterative training on the Q growth rate prediction paths. First, the generator attempts to predict the growth rate using the genotype data in the current training set. This process continuously adjusts the parameters of the generator so that its output growth rate is more consistent with the real data. After the generator generates the predicted growth rate, the discriminator compares the generated growth rate with the real growth rate data to determine its authenticity. The goal of the discriminator is to accurately distinguish between the real and generated growth rate data. The generator and the discriminator are updated alternately until the generator can generate accurate growth rate predictions and the discriminator can effectively distinguish between the real and predicted data. The training process continues to iterate until the gap between the growth rate generated by the generator and the real growth rate is small enough and the discriminator can make a correct judgment. The training is then stopped to obtain the Q growth rate prediction paths that have been trained, and the growth rate prediction channel is obtained based on the combination of the Q growth rate prediction paths.

[0045] Then, the breeding quantity coefficient is multiplied by Q and rounded to obtain the growth rate prediction accuracy P, where P is a positive integer less than or equal to Q; then, P growth rate prediction paths are randomly selected from the Q growth rate prediction paths, and the genotype data are input into the randomly selected P growth rate prediction paths, and the prediction output obtains P predicted growth rates, and the mean of the P predicted growth rates is calculated to obtain the growth rate.

[0046] S400: Matching and identifying the multiple meat pattern images and growth rates to obtain multiple matching degrees, and calculating and obtaining evaluation accuracy based on the meat pattern image generation accuracy and growth rate prediction accuracy.

[0047] Furthermore, step S400 of the present invention further includes:

[0048] S410: Based on the fat distribution in the multiple meat pattern patterns, multiple fat uniformity parameters are calculated; S420: The multiple fat uniformity parameters are input into the meat growth rate mapper to map and obtain multiple mapping growth rates, wherein the meat growth rate mapper includes a mapping relationship between sample fat uniformity parameters and sample growth rates; S430: Based on the deviations between the multiple mapping growth rates and the growth rates, multiple matching degrees are calculated, and matching accuracy is calculated; S440: Based on the path accuracy of the meat pattern pattern generation path and the growth rate prediction path selected according to the meat pattern pattern generation accuracy and the growth rate prediction accuracy, basic accuracy is calculated; S450: Based on the matching accuracy and basic accuracy, evaluation accuracy is calculated.

[0049] Specifically, fat uniformity is one of the important indicators for evaluating the quality of meat patterns (such as snowflake patterns). The uniformity of fat distribution directly affects the taste and tenderness of the meat. Fat uniformity analysis is performed based on the fat distribution within the multiple meat pattern patterns. First, a first meat pattern pattern is randomly selected, and multiple fat areas are randomly selected from the first meat pattern pattern for measurement. These fat areas can be local areas at different locations to ensure the representativeness and diversity of the sample. Then, in each selected fat area, the fat width is measured (for example, a standard width is selected, such as 2mm), and the fat width data at that location is recorded. The width of the fat area can be automatically detected by an image processing tool or manually calibrated. Furthermore, for each fat area in the meat pattern pattern, multiple measurement values ​​are collected, and the variance of the fat width in the multiple measurement values ​​is calculated. The variance is an indicator that measures the degree of dispersion of a set of data distribution. The inverse of the variance is set as the first fat uniformity parameter, and multiple fat uniformity parameters are analyzed in sequence. The fat uniformity parameter is inversely proportional to the variance. The smaller the variance, the more uniform the fat distribution.

[0050] Growth rate not only affects the weight and muscle development of beef cattle, but also directly affects the distribution and formation of fat, thereby affecting the final meat texture pattern. For example, when beef cattle grow faster, muscle tissue growth usually takes precedence over fat accumulation. Excessive growth may cause muscle tissue to overexpand in a short period of time, failing to provide sufficient time and space for the even distribution of fat. In contrast, slower-growing beef cattle have more time to evenly distribute fat, especially when the muscles are already mature. Fat can be more evenly filled between muscle fibers, forming a more uniform snowflake pattern. First, a set of sample fat uniformity parameters and a set of sample growth rates are collected, where the sample fat uniformity parameters and the sample growth rates correspond one to one. Then, a meat texture growth rate mapper is constructed based on the sample fat uniformity parameter set and the sample growth rate set. For example, a regression model (such as linear regression, nonlinear regression, etc.) is used to fit the relationship between the fat uniformity parameter and growth rate. The regression model can help analyze the impact of fat distribution uniformity on growth rate. The model is trained using the sample fat uniformity parameter set and the sample growth rate set to obtain a meat texture growth rate mapper. The plurality of fat uniformity parameters are further input into a meat growth rate mapper, and mapped to obtain a plurality of mapped growth rates.

[0051] Then, the deviations of the multiple mapped growth rates and the growth rates are calculated respectively to obtain multiple growth rate deviations, wherein the growth rate deviation is the absolute value of the difference between the mapped growth rate and the growth rate; then, the ratio of the growth rate deviation to the growth rate is set as the deviation percentage, and the matching degree is obtained by subtracting the deviation percentage from 1, and multiple matching degrees of the multiple growth rate deviations are calculated, and the multiple matching degrees are averaged to obtain the matching accuracy.

[0052] Obtain P meat pattern prediction accuracy rates for the P meat pattern generation paths selected for the meat pattern generation accuracy, and average the P meat pattern prediction accuracy rates to obtain a mean meat pattern prediction accuracy rate. Optionally, the P meat pattern generation paths selected in the aforementioned content may be tested to obtain P meat pattern prediction accuracy rates.

[0053] Secondly, the P growth prediction accuracy rates for the P growth rate prediction paths are obtained, and the mean growth prediction accuracy rate is calculated. Finally, the mean of the pattern prediction accuracy rate and the mean of the growth prediction accuracy rate are averaged to obtain the basic accuracy. The matching accuracy and the basic accuracy are then averaged, and the average calculation result is set as the evaluation accuracy. The evaluation accuracy is a comprehensive indicator obtained by comprehensively considering multiple factors such as meat pattern generation, fat distribution uniformity, and growth rate prediction. It can effectively reflect the predictive ability and accuracy of the model, thereby providing a more scientific and accurate basis for the breeding value assessment of beef cattle.

[0054] S500: Evaluate and calculate the breeding value based on the multiple flesh pattern diagrams and growth rates, compensate the breeding value using the evaluation accuracy, and obtain a breeding value interval as a breeding value analysis result.

[0055] Furthermore, step S500 of the present invention further includes:

[0056] S510: Calculate the ratio of the growth rate to multiple mapped growth rates, and calculate the mean to obtain the growth rate breeding value; S520: Calculate the ratio of the multiple fat uniformity parameters to the preset fat uniformity parameters to obtain the fat breeding value; S530: Perform weighted calculation on the fat breeding value and the growth rate breeding value to obtain the breeding value.

[0057] Specifically, the ratio of the growth rate and multiple mapped growth rates is calculated to obtain multiple growth rate ratios, and the mean of the multiple growth rate ratios is calculated to obtain the growth rate breeding value, wherein the larger the growth rate breeding value, the closer the growth rate is to the ideal value, and the faster the growth rate is, the better the breeding effect is; on the other hand, the ratios of the multiple fat uniformity parameters and the preset fat uniformity parameters are calculated respectively, and the fat breeding value is obtained after mean calculation, wherein the larger the fat breeding value, the closer the fat uniformity is to the ideal value, the more uniform the fat is, and the better the breeding effect is.

[0058] Then, weights are assigned to the fat breeding value and the growth rate breeding value. The weight values ​​can be set according to the impact of the indicators on the overall breeding effect. The greater the impact, the greater the corresponding weight. This can be determined through historical data analysis or expert evaluation. For example, if more attention is paid to growth rate, a larger weight can be assigned to the growth rate breeding value; if more attention is paid to meat quality, the weight of the fat breeding value can be increased; for example, the weights of the fat breeding value and the growth rate breeding value are 0.6 and 0.4 respectively.

[0059] Then, the fat breeding value and growth rate breeding value are weighted and calculated according to the weight configuration result to obtain the final breeding value. The breeding value reflects the comprehensive performance of each beef cattle in terms of growth rate and fat distribution uniformity. The higher the breeding value, the closer the performance of the beef cattle in these two aspects is to the ideal state, and the better the breeding effect.

[0060] Furthermore, step S500 of the present invention further includes:

[0061] S540: Calculate and obtain a compensation coefficient based on the evaluation accuracy; S550: Use the compensation coefficient to perform interval compensation on the breeding value to obtain a breeding value interval as a breeding value analysis result.

[0062] Specifically, the sum of the evaluation accuracy plus 1 is used as the upper limit of the compensation coefficient, and the difference between the two is used as the lower limit of the compensation coefficient. Finally, the upper and lower limits of the compensation coefficient are multiplied by the breeding value, respectively, to perform interval compensation on the breeding value to obtain the breeding value interval. For example, assuming an evaluation accuracy of 0.05 and a breeding value of 0.6, the breeding value interval is (1-0.05)*0.6 to (1+0.05)*0.6, that is, 0.57 to 0.63. Finally, the breeding value interval is used as the breeding value analysis result. By performing interval compensation on the breeding value based on the evaluation accuracy, the comprehensiveness, rationality, and reliability of the breeding value analysis results can be improved.

[0063] The embodiment of the present invention provides a method for evaluating and analyzing the genomic breeding values ​​of multiple beef cattle breeds, which has at least the following technical effects:

[0064] The target beef cattle are sequenced using a gene chip to obtain genotype data of the target beef cattle and the beef cattle breed of the target beef cattle; then, the meat texture pattern generation accuracy is configured according to the beef cattle breed, and multiple meat texture pattern patterns are generated based on the genotype data and the meat texture pattern generation accuracy; on the other hand, the growth rate prediction accuracy is configured according to the beef cattle breed, and the growth rate phenotype is predicted based on the genotype data to predict the growth rate; the multiple meat texture pattern patterns and growth rates are further matched and identified to obtain multiple matching degrees, and the evaluation accuracy is calculated based on the meat texture pattern generation accuracy and the growth rate prediction accuracy; then, the breeding value is evaluated and calculated based on the multiple meat texture pattern patterns and growth rate; finally, the breeding value is compensated using the evaluation accuracy to obtain a breeding value range as the breeding value analysis result; that is, by performing multi-dimensional analysis based on the meat texture pattern and the growth rate of beef cattle, and combining it with machine learning and other methods for intelligent evaluation, the breeding potential of beef cattle can be evaluated more comprehensively and scientifically, and the scientificity, accuracy and reliability of the beef cattle breeding value evaluation can be significantly improved.

[0065] Although preferred embodiments of the present invention have been described, additional changes and modifications to these embodiments may occur to those skilled in the art once the basic inventive concepts become known.

[0066] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for evaluating and analyzing the genomic breeding values ​​of multiple beef cattle breeds, characterized in that: Methods include: Using a gene chip, obtaining genotype data of a target beef cattle and obtaining the beef cattle breed of the target beef cattle; According to the beef cattle breed, the meat pattern map generation accuracy is configured, and according to the genotype data and the meat pattern map generation accuracy, a plurality of meat pattern maps are generated, including: Obtain the total number of beef cattle raised, and obtain the number of beef cattle breeds raised; Calculate the ratio of the breeding quantity to the total breeding quantity to obtain the breeding quantity coefficient; The meat pattern image generation path of the training integration number Q is obtained to obtain the meat pattern image generation channel, where Q is a positive integer, including: Based on the beef quality test data in the historical period, the sample genotype data set and the sample meat pattern map set are collected; Randomly collecting Q copies of the sample genotype data set and the sample meat texture pattern map set with replacement to generate supervised training data; Based on the generative adversarial network, Q meat pattern generation paths are constructed, each of which includes a generator and a discriminator. The Q pieces of supervised training data are respectively used to perform alternating iterative training on the generators and discriminators in the Q meat pattern image generation paths until the training is completed, and a meat pattern image generation channel is obtained by combining them; According to the breeding quantity coefficient and Q, the meat pattern generation accuracy P is calculated, where P is a positive integer less than or equal to Q; Inputting the genotype data into P randomly selected meat pattern generation paths to generate P meat pattern maps; According to the beef cattle breed, the growth rate prediction accuracy is configured, and the growth rate phenotype is predicted based on the genotype data to predict the growth rate, including: Obtain the total number of beef cattle raised, as well as the number of beef cattle breeds raised, and calculate the breeding number coefficient; Training an integrated number Q of growth rate prediction paths to obtain growth rate prediction channels, wherein the input data in the training process of each growth rate prediction path is genotype data and the supervision data is growth rate; According to the breeding quantity coefficient and Q, the growth rate prediction accuracy P is calculated, where P is a positive integer less than or equal to Q; Inputting the genotype data into randomly selected P growth rate prediction paths, predicting output to obtain P predicted growth rates, and calculating the mean to obtain the growth rate; Matching and identifying the plurality of meat pattern images and growth rates to obtain a plurality of matching degrees, and calculating and obtaining an evaluation accuracy by combining the meat pattern image generation accuracy and the growth rate prediction accuracy; According to the multiple flesh pattern diagrams and growth rates, the breeding value is evaluated and calculated, and the breeding value is compensated using the evaluation accuracy to obtain a breeding value interval as a breeding value analysis result.

2. The method for evaluating and analyzing the genomic breeding values ​​of multiple varieties of beef cattle according to claim 1, characterized in that: Using a gene chip to obtain genotype data of a target beef cattle and obtain the breed of the target beef cattle, including: Gene chips were used to sequence the target beef cattle to obtain genotype data; Obtain the target beef cattle breed.

3. The method for evaluating and analyzing the genomic breeding values ​​of multiple varieties of beef cattle according to claim 1, characterized in that: Matching and identifying the multiple meat texture patterns and growth rates to obtain multiple matching degrees, and calculating and obtaining an evaluation accuracy based on the meat texture pattern generation accuracy and growth rate prediction accuracy, including: Calculating and obtaining a plurality of fat uniformity parameters according to the fat distribution in the plurality of meat texture pattern images; Inputting the plurality of fat uniformity parameters into a meat quality growth rate mapper to map and obtain a plurality of mapped growth rates, wherein the meat quality growth rate mapper includes a mapping relationship between the sample fat uniformity parameters and the sample growth rates; Calculating a plurality of matching degrees according to the plurality of mapping growth rates and deviations of the growth rates, and calculating a matching accuracy; Calculate the basic accuracy based on the path accuracy of the meat pattern generation path and the growth rate prediction path selected according to the meat pattern generation accuracy and the growth rate prediction accuracy; The evaluation accuracy is calculated based on the matching accuracy and the basic accuracy.

4. The method for evaluating and analyzing the genomic breeding values ​​of multiple varieties of beef cattle according to claim 3, characterized in that: Based on the plurality of meat pattern diagrams and growth rates, the breeding values ​​are evaluated and calculated, including: Calculating a ratio of the growth rate to a plurality of mapped growth rates, and calculating an average to obtain a growth rate breeding value; Calculating a ratio of the plurality of fat uniformity parameters to a preset fat uniformity parameter to obtain a fat breeding value; The fat breeding value and the growth rate breeding value are weighted and calculated to obtain the breeding value.

5. The method for evaluating and analyzing the genomic breeding values ​​of multiple varieties of beef cattle according to claim 4, characterized in that: The breeding value is compensated using the evaluation accuracy to obtain a breeding value interval as a breeding value analysis result, including: Calculating a compensation coefficient based on the evaluation accuracy; The compensation coefficient is used to perform interval compensation on the breeding value to obtain a breeding value interval as a breeding value analysis result.

Images