Molecular identity card for identifying variety of Tianhua mutton sheep and application of molecular identity card

Abstract

The invention discloses a molecular identity card for identifying the variety of Tianhua mutton sheep and application of the molecular identity card, the molecular identity card is a specific SNP site combination for identifying the variety of Tianhua mutton sheep, the combination is composed of 36 SNP sites, is located in a sheep reference genome AR-UIRamv2.0, has high specificity and machine learning model classifier accuracy, and can be used for identifying the variety of Tianhua mutton sheep. The method can be used for identifying Tianhua mutton sheep, some varieties similar to the Tianhua mutton sheep in appearance and other genetic resource varieties of Gansu Province sheep. According to the invention, variety-specific molecular markers are screened at a genome level through a bioinformatics technology, genetic differences between Tianhua mutton sheep and some varieties with similar appearances and other genetic resource varieties of Gansu Province sheep can be revealed, and the Tianhua mutton sheep can be identified through a machine learning classifier and genetic evolutionary tree analysis according to characteristic SNP (Single Nucleotide Polymorphism). Guarantee is provided for fighting against counterfeit and shoddy varieties and reasonably protecting and utilizing variety resources of Tianhua mutton sheep.

Description

Technical Field

[0001] The invention belongs to the technical field of sheep breed identification, and relates to a molecular ID card for identifying Tianhua mutton sheep breeds. The invention also relates to the application of the molecular ID card in the identification of Tianhua mutton sheep germplasm resources. Background Art

[0002] Tianhua mutton sheep is a new breed of fine-wool mutton sheep that can adapt to the high-altitude climate. It was bred over 15 years with imported South African mutton Merino sheep as the father and Gansu alpine fine-wool sheep as the mother. This breed has the advantages of good meat and wool performance, high fertility, and strong adaptability. It conforms to the current market demand for the transition from wool sheep to both meat and wool sheep, and at the same time has the maternal adaptability to the local high-altitude cold environment. The breeding of this breed solves the shortcomings of the original breeds in the high-altitude and surrounding areas, such as slow growth, low fertility, long market cycle, poor meat performance, and degraded wool performance. It can effectively ensure that the precious fine-wool sheep resources will not be mixed and destroyed due to fluctuations in the wool market and the gap in the comparative benefits of wool and mutton, meet the important needs of high-altitude and cold areas, especially fine-wool sheep producing areas, to maintain wool and increase meat, and provide a good source guarantee for the upgrading of sheep breeds in such areas.

[0003] At present, the identification of sheep breeds mostly relies on appearance identification or blood pedigree analysis. This type of identification method is not only time-consuming and laborious, but also easily interfered by subjective factors or environmental factors, making it difficult to achieve accurate identification. Today, genomics is developing rapidly, and molecular marker technology has also made great progress in the field of genetic breeding. The new generation of molecular marker SNPs has the characteristics of large number, wide distribution, genetic stability, high specificity, and easy standardization and automated detection. By selecting signals combined with machine learning classifiers, the breed-specific SNPs screened out can achieve accurate identification of sheep breeds, quickly and efficiently. At present, there is no SNP molecular marker technology for the identification of Tianhua sheep breeds. Summary of the invention

[0004] The purpose of the present invention is to provide a molecular identity card for identifying the Tianhua mutton breed, which is used for the identification of Tianhua mutton germplasm resources, is conducive to the further promotion and breeding of its breed, and prevents the loss of genetic resources or economic losses of herders due to breed impersonation or hybrid substitution, thereby promoting the long-term development of Tianhua mutton sheep in the sheep breeding industry.

[0005] To achieve its purpose, the present invention adopts the following technical solution:

[0006] The present invention provides a molecular ID card for identifying the Tianhua mutton breed, wherein the molecular ID card consists of 36 characteristic SNP sites located in the sheep reference genome ARS-UI_Ramb_v2.0, and the SNP sites are a set of Tianhua mutton-specific SNP sites screened by selection signals and maximum correlation minimum redundancy (mRMR), and the specific information is shown in Table 1 below.

[0007] Table 1.

[0008]

[0009]

[0010] The above-mentioned Tianhua mutton molecular ID card can be used for the identification of Tianhua mutton germplasm resources. The application method is as follows:

[0011] (1) Perform whole genome resequencing on the unknown breed sheep sample to obtain original genome data;

[0012] (2) comparing the raw data obtained in step (1) with the sheep reference genome ARS-UI_Ramb_v2.0 and performing genotyping to obtain the genotype data of the sheep to be tested;

[0013] (3) filtering the genotype data of step (2), and then self-filling it through Beagle software, and then extracting the SNP sites corresponding to the molecular ID card of Tianhua sheep;

[0014] (4) The genotype data of the SNP sites on the molecular ID card of the Tianhua mutton population are merged, and a genetic evolution tree is constructed to determine whether it is a Tianhua mutton mutton genetic resource; or its genotype format is converted into 012 format and input into a trained python machine learning classifier model, and its predict function is used to determine whether it is a Tianhua mutton mutton genetic resource.

[0015] The beneficial effects of the present invention are:

[0016] The present invention obtains the original genome data of 55 Tianhua mutton sheep, 11 South African mutton merino sheep and 11 Gansu alpine fine wool sheep through whole genome resequencing, and compares and analyzes the whole genome genotype data of 117 individuals with 7 other sheep genetic resources similar to Tianhua mutton in appearance or other sheep in Gansu Province, and screens out 36 specific SNP combinations of Tianhua mutton sheep as Tianhua mutton sheep specific molecular ID cards. The Tianhua mutton sheep specific molecular ID card of the present invention has obvious Tianhua mutton breed specificity. For the 36 SNPs, the genetic evolution tree can be constructed to clearly distinguish Tianhua mutton sheep from non-Tianhua mutton sheep groups. The machine learning classifier model can also be used to predict the breed of the test set, and the accuracy, precision and F1 scores of the support vector machine, random forest, naive Bayes and logistic regression models used have reached more than 95%. This shows that this molecular ID card can use less genotypic information to simply and clearly distinguish between Tianhua mutton sheep populations and non-Tianhua mutton sheep populations in Gansu Province and neighboring areas, providing a scientific and powerful basis for the identification and protection of Tianhua mutton sheep germplasm resources, and laying the foundation for the further improvement of Tianhua mutton sheep and its promotion in the sheep farming industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a phylogenetic tree diagram of 10 sheep breeds;

[0018] Figure 2 The accuracy of the feature SNPs obtained for the selection signal in the machine learning classifier;

[0019] Figure 3 The accuracy of the machine learning classifier for different numbers of SNPs screened for mRMR between 10 and 100;

[0020] Figure 4 The accuracy of the machine learning classifier for different numbers of SNPs between 30 and 40 screened for mRMR;

[0021] Figure 5 This is a phylogenetic tree diagram constructed based on 36 Tianhua sheep molecular ID SNPs;

[0022] Figure 6 This is a diagram showing the results of genetic evolution analysis of an application example of the present invention. DETAILED DESCRIPTION

[0023] In order to better illustrate the purpose, technical solutions and advantages of the present invention, the present application will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0024] The Tianhua mutton described in the present invention comes from the Tianhua mutton core breeding farm of Gansu Lantian Tonghe Agriculture Co., Ltd. in Tianzhu Tibetan Autonomous County, Wuwei City, Gansu Province.

[0025] Example 1

[0026] A method for screening specific molecular ID cards of Tianhua mutton germplasm resources, comprising the following steps: 1. Sample data acquisition

[0027] Blood samples were collected from 55 Tianhua sheep, 11 South African merino sheep and 11 Gansu alpine fine wool sheep. Genomic DNA was extracted using the phenol-chloroform method and then sent to Beijing Berry Genomics Biotech Co., Ltd. for DNA library construction. Illumina PE150 strategy was used for sequencing to obtain the fastq.gz data of each sample. In addition, fastq.gz data of 119 sheep germplasm resources in Gansu Province and breeds similar to Tianhua sheep in appearance, including Australian Merino sheep (n=7), Chinese Merino sheep (n=37), German Merino sheep (n=17), Lanzhou big-tailed sheep (n=4), Minxian black fur sheep (n=9), Oula sheep (n=10) and Tan sheep (n=35) were collected from the NCBI database.

[0028] 2. Data Format Conversion

[0029] BaseNumber DNA sequencing data analysis software was used for quality control, alignment and SNP detection. The reference genome version for alignment was the sheep reference genome ARS-UI_Ramb_v2.0, thus obtaining a total of 196 samples in vcf.gz format files.

[0030] 3. Site Filtering

[0031] The vcf.gz file was filtered using bcftools and plink software, autosomal and biallelic SNPs were retained, sites with minor allele frequency (MAF) less than 0.05 were filtered out, and all sites with deletions were removed so that they could be used as features for subsequent machine learning classifiers. Plink software was used to remove linked sites, and the parameters set were: window size 50 SNPs, step size 10 SNPs, and LD threshold set to 0.2.

[0032] 4. Population genetic analysis and selection of characteristic SNPs

[0033] The filtered sites were used to construct the genetic evolutionary tree for all samples, and the genetic distance matrix was constructed using plink software. The evolutionary tree was then constructed using MEGA software and beautified using the iTOL website. Figure 1As shown, the 10 selected breeds are well clustered and have obvious differentiation. The characteristic SNP selection adopts a strategy combining two selection signals. First, the Pi values ​​(nucleotide diversity) of the 10 breeds were calculated using the vcftools software, and the window size was set to 10,000. Then, the logarithm of the Pi ratio between Tianhua sheep and the other 9 breeds was calculated using a self-written script, i.e., the ln-Pi ratio. The windows with the top 2% of the ln-Pi ratio values ​​between each breed pair, i.e., the top 1% of the windows and the bottom 1% of the windows, were obtained, and the SNPs in the above windows were extracted using bcftools. Secondly, the genetic differentiation index (Fst) of each site in the aforementioned window between Tianhua sheep and the other breeds was calculated using the vcftools software, and the top 1% of the SNP sites of each breed pair were extracted. After the intersection of these sites, they were used as the characteristic SNP sites of Tianhua sheep.

[0034] 5. Testing of characteristic SNPs

[0035] Extract the SNP sites obtained in the previous step, and use plink software to convert the SNP site information into 012 encoding format, where 0 represents homozygous reference genome sites, 1 represents heterozygous, and 2 represents homozygous mutation sites. Take each SNP site information as a feature, and use the support vector machine model, random forest model, naive Bayes model and logistic regression model in the sklearn library of python to train the classifier model. The labels of all samples are divided into Tianhua sheep group and non-Tianhua sheep group, 0 represents Tianhua sheep, 1 represents non-Tianhua sheep, and 80% of the samples are used as training sets to train the classification model, and the remaining 20% ​​are used as test sets to test the accuracy of the model. The model training uses parameter grid search and 5-fold cross-validation to obtain the optimal model. The training set test results are as follows. Figure 2 As shown: It shows that the 164 SNP loci obtained in step 4 can well distinguish Tianhua sheep from non-Tianhua sheep.

[0036] 6. Reduction of characteristic SNPs

[0037] The SNPs obtained in the previous step are further screened, and the maximum correlation and minimum redundancy (mRMR) algorithm is used to find a set of features in the original feature set that are most correlated with the final output result, but have the least correlation with each other. This is done using the python mrmr library, with different numbers of SNPs set to observe the accuracy of the machine learning model classifier under different SNPs. While ensuring good accuracy, try to select a smaller number of SNP site combinations. Compare the accuracy, precision, and F1 scores of the four machine learning classifiers with different numbers of SNPs to obtain the most appropriate combination of SNP numbers. The three evaluation indicators of SNPs under different numbers are as follows: Figure 3 and4 As shown, 36 SNPs were finally selected as the final Tianhua sheep-specific molecular identity card.

[0038] Example 2

[0039] Method 1 for identifying Tianhua mutton germplasm resources using Tianhua mutton molecular ID card, the steps are:

[0040] 1. Acquisition of SNP site data of the molecular ID card of the sheep to be tested

[0041] Blood samples or tissue samples were collected from the sheep to be tested, DNA was extracted, and sent to the sequencing company for sequencing. For the data after sequencing, the vcf.gz file was obtained by genome alignment and genotyping using BaseNumber DNA sequencing data analysis software or BWA+GATK. The bcftools software was further used to extract the SNP sites of the Tianhua sheep molecular ID card according to the position information. For missing sites, the genotype filling software Beagle was used for self-filling.

[0042] 2. SNP site genotype format conversion

[0043] For the filled vcf.gz file, plink software was used to convert the SNP site information into 012 encoding format, where 0 represents homozygous reference genome site, 1 represents heterozygous, and 2 represents homozygous mutation site. Ensure that the final characteristic SNP is consistent with the Tianhua sheep molecular ID card.

[0044] 3. Machine Learning Classifier Prediction

[0045] Machine learning classifier prediction relies on python. First, numpy and pandas are used for data processing, and joblib is used to save and load models and other objects. Sklearn is used to load the classifier model, and the machine learning model is used to determine whether it is a Tianhua mutton genetic resource based on its predict function. If the output result is 0, it is a Tianhua mutton, and if the output is 1, it is not a Tianhua mutton.

[0046] Example 3

[0047] Method 2 for identifying Tianhua mutton germplasm resources using Tianhua mutton molecular ID card, the steps are:

[0048] 1. Acquisition of SNP site data of the molecular ID card of the sheep to be tested

[0049] Blood samples or tissue samples are collected from the sheep to be tested, DNA is extracted, and sent to a sequencing company for sequencing. For the data after sequencing, the BaseNumber DNA sequencing data analysis software or BWA+GATK is used for genome alignment and genotyping to obtain a vcf.gz file.

[0050] 2. SNP site data merging and evolutionary tree construction

[0051] Use bcftools software to merge the vcf.gz file of the sheep to be tested with the existing vcf.gz file of Tianhua sheep, and extract the 36 Tianhua sheep molecular ID loci in the data. Use plink software to calculate the genetic distance matrix, and MEGA software to construct the genetic evolutionary tree. When it is detected that the genetic distance of the sheep to be tested is close to the genetic distance of the Tianhua sheep population, it can be determined that the sheep to be tested is Tianhua sheep.

[0052] Application Examples

[0053] The application of Tianhua mutton molecular ID card in identifying Tianhua mutton germplasm resources is as follows:

[0054] 1. Select 81 new Tianhua sheep as the sheep to be tested, collect blood samples from the sheep to be tested, extract DNA and send them to Boridi Biotechnology Co., Ltd. for sequencing. For the sequencing data returned by the sequencing company, use BaseNumberDNA sequencing data analysis software to obtain the vcf.gz file, and the reference genome used is ARS-UI_Ramb_v2.0. Further use bcftools software to extract 36 specific SNP sites of the Tianhua sheep molecular ID card according to the position information, and merge with the existing population used to construct the Tianhua sheep molecular ID card.

[0055] 2. Use plink software to calculate the genetic distance matrix, MEGA software to construct the genetic evolution tree, and use the iTOL website to beautify the genetic evolution tree. The results of genetic evolution analysis are as follows: Figure 6 As shown, it can be seen that the sheep to be tested (i.e., the newly collected Tianhua mutton sheep) have a close genetic relationship with the original Tianhua mutton sheep, and it can be determined that the sheep to be tested belong to the Tianhua mutton sheep germplasm genetic resources.

[0056] 3. For the vcf.gz file obtained in the first step, the missing sites are filled using the genotype filling software Beagle. Then the SNP site information is converted into 012 encoding format using plink software, where 0 represents homozygous reference genome sites, 1 represents heterozygous, and 2 represents homozygous mutation sites. Use python's numpy and pandas to process the original data, and use joblib to save and load data and models. Use sklearn to load the classifier model. Taking the support vector machine with the best performance in the training process as an example, the predict and predict_proba functions of the trainer model are used to judge the sheep to be tested. The model output results are shown in Table 2.

[0057] Table 2.

[0058]

[0059]

[0060]

[0061] Among them, in Table 1, the output result is 0 for Tianhua sheep, and the output is 1 for non-Tianhua sheep. The output results of the newly collected 81 sheep to be tested are all 0, and the prediction accuracy reaches more than 90%. It can be judged that the newly collected sheep to be tested are all Tianhua sheep.

[0062] The above-described embodiments only express several implementation methods of the present invention. It should be pointed out that, for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A molecular ID card for identifying the Tianhua mutton breed, characterized in that: The molecular ID card consists of 36 characteristic SNP sites located in the sheep reference genome ARS-UI_Ramb_v2.0, and the SNP sites include ID1-ID36, wherein: ID1-ID4 are located at positions 96522898, 139740899, 153767171, and 213201360 on chromosome 1, respectively; ID5-ID6 are located at positions 71425831 and 210125476 on chromosome 2, respectively; ID7-ID10 are located at positions 49175961, 118709694, 132448684, and 218813552 on chromosome 3, respectively; ID11-ID16 are located on chromosome 4 at positions 77459709, 80420769, 90323194, 92536278, 104178591, and 116757608, respectively; ID17-ID18 are located at positions 83,142,360 and 95,932,782 on chromosome 5, respectively; ID19 is located at position 66431083 on chromosome 6; ID20-21 are located at positions 20,949,969 and 98,180,525 on chromosome 7, respectively; ID22 is located at position 42520240 on chromosome 8; ID23-24 are located at positions 34397831 and 40091099 on chromosome 9, respectively; ID25-27 are located at positions 25822711, 33041271, and 42569697 on chromosome 14, respectively; ID28-29 are located at positions 7988627 and 15198939 on chromosome 15, respectively; ID30 is located at position 24325191 on chromosome 16; ID31 is located at position 19258771 on chromosome 17; ID32 is located at position 21749026 on chromosome 18; ID33 is located at position 37662890 on chromosome 20; ID34 is located at position 16097845 on chromosome 21; ID35-ID36 are located at position 37701101 and 39398635 of chromosome 25.

2. A molecular ID card for identifying the Tianhua sheep breed according to claim 1, characterized in that: The mutation types of the ID1, ID4, ID6, ID8, ID13, ID14, ID23, ID29, and ID30 sites are G / A, the mutation types of the ID2, ID3, ID27, and ID28 sites are T / G, the mutation types of the ID5, ID10, ID16, ID17, ID19, ID20, ID21, ID24, ID25, and ID33 sites are C / T, the mutation type of the ID7 site is G / T, the mutation type of the ID9 and ID32 sites is A / G, the mutation type of the ID11 and ID34 sites is T / A, the mutation type of the ID12 site is T / C, the mutation type of the ID 15 and ID22 sites is C / A, the mutation type of the ID18, ID26, ID35, and ID36 sites is T / C, and the mutation type of the ID31 site is A / T.

3. Application of the molecular ID card according to claim 1 or 2 in identification of Tianhua mutton germplasm resources.

4. The application of the molecular ID card in the identification of Tianhua mutton germplasm resources according to claim 3 is characterized in that: The steps include: (1) Perform whole genome resequencing on the unknown breed sheep sample to obtain original genome data; (2) comparing the raw data obtained in step (1) with the sheep reference genome ARS-UI_Ramb_v2.0 and performing genotyping to obtain the genotype data of the sheep to be tested; (3) filtering the genotype data of step (2), and then self-filling it through Beagle software, and then extracting the SNP sites corresponding to the molecular ID card of Tianhua sheep; (4) The genotype data of the SNP sites on the molecular ID card of the Tianhua mutton population are merged, and a genetic evolution tree is constructed to determine whether it is a Tianhua mutton mutton genetic resource; or its genotype format is converted into 012 format and input into a trained python machine learning classifier model, and its predict function is used to determine whether it is a Tianhua mutton mutton genetic resource.

Images