CN115341044A - A method for predicting daily weight gain of pigs using microbes and their associated SNP sites - Google Patents

Abstract

The invention discloses a method for predicting daily gain of pigs by utilizing microorganisms and related SNP sites thereof, and relates to the technical field of pig genetic genes, wherein the microorganisms comprise terrispobacter petrilerolearius and Prevotella copri. The method comprises the following steps: sample collection, microorganism DNA extraction and 16S rDNA sequencing, data quality control, correlation analysis and verification of the influence of the microorganism and related SNP loci on the average daily gain of pigs. The method of the invention can well predict the daily gain of the pig, effectively and accurately breed the pig breed with high daily gain, and is beneficial to achieving the purpose of improving the production level and the economic benefit of the pig raising industry.

Description

A method for predicting daily weight gain of pigs using microbes and their related SNP sites

technical field

The invention relates to the technical field of pig genetics, in particular to a method for predicting daily weight gain of pigs by using microorganisms and related SNP sites.

Background technique

my country is a large pig-raising country, and the market demand for pork production and quality is increasing day by day. Increasing pork production and improving pork carcass quality have become the work that breeding scientists have been exploring for a long time. The early breeding work mainly focused on the phenotypic selection of pigs. With the continuous advancement of genomic work and the use of molecular markers, breeding selection through single nucleotide polymorphism (single nucleotide polymorphism, SNP) markers has become the current trend. Mainstream way. However, with new breakthroughs in the research of gut microbes in recent years, the importance of gut microbes is well known, and its influence on the phenotype of pigs cannot be ignored.

There is a large number of microbiota in the gastrointestinal tract of mammals, and the number of genes is about 1 to 1.3 times that of the host gene. The role of the microbiota in the gut and their metabolites cannot be ignored. More and more studies have shown that the gut can affect the host's immunity and nutrient metabolism through the gut-brain axis and the gut-liver axis. In livestock and poultry, some scholars have found that gut microbes not only affect the phenotype of the host, such as growth traits, meat quality traits, etc., but also that gut microbes are heritable. However, there is still little research on the inclusion of gut microbes into livestock breeding and the precise selection of livestock.

With the continuous development of high-throughput technology, 16S sequencing technology provides technical support for the research of intestinal microorganisms. The 16S rDNA gene exists in the genomes of all bacteria and is highly conserved. 16S sequencing is to extract the total DNA from the sample, amplify and sequence the target fragment region (that is, a partial region of 16S rDNA), and then obtain the microbial community information of the sample species through data analysis. At present, the sequencing platforms used on the market mainly include the second-generation sequencing platform IlluminaMiseq/HiSeq. Due to the limitation of read length, 16S sequencing can only select single-V region, double-V region or triple-V region as the target fragment region, and because V3, V4, V5 The specificity is better, and the database information is more complete. Among them, the primers 341F and 806R were selected for the V3-V4 region. The primers have a high coverage rate in bacteria and archaea, and can detect the distribution of bacteria and archaea at the same time. The V4-V5 region has good specificity and complete database information, which is the best choice for bacterial diversity analysis and annotation.

To sum up, after massive searches by applicants, there are at least traditional artificial breeding or breeding methods based on genome information in this field, and the marginal benefits are constantly decreasing. New technologies are needed to break through the current breeding margins and carry out precise breeding. Pigs with more significant daily gain are selected. Therefore, it is necessary to develop or improve a method for predicting daily gain of pigs using microorganisms and their related SNP sites.

Contents of the invention

Based on this, in order to solve the problem that the marginal benefits of traditional artificial breeding or breeding methods based on genome information are constantly decreasing, new technologies are needed to break through the current breeding margin, carry out precise breeding, and select pigs with more significant daily weight gain. , the present invention provides a method for predicting daily weight gain of pigs using microorganisms and their associated SNP sites, the specific technical scheme is as follows:

A method for predicting daily gain of pigs using microorganisms and their associated SNP sites, the microorganisms including Terrisporobacter petroleumarius and Prevotella copri.

Further, the SNP sites related to the Terrisporobacter petroleumarius include rs339933029, rs333900969, rs332402643, rs338935223, rs80986577 and rs81415286.

Further, the gene related to the SNP site rs332402643 is PDZRN4, the gene related to the SNP site rs338935223 is WWOX, and the gene related to the SNP site rs80986577 is RAD51B.

Further, the SNP sites related to the Prevotella copri include rs81437804 and rs343769713.

Further, the gene related to the SNP site rs343769713 is IRS1.

Further, the method includes the following steps:

Sample collection, microbial DNA extraction and 16S rDNA sequencing, data quality control, correlation analysis and verification of the effects of microorganisms and their associated SNP sites on the average daily gain of pigs.

Further, the sample collection includes the following steps:

When the pigs are 64-150 days old, use the performance automatic measurement system to measure the average daily gain of pigs, and when the body weight reaches 130±5KG, the measurement ends; after the original body weight data is quality controlled, calculate the average daily gain of each individual;

Pig ear tissue was collected for DNA extraction, and the pigs were genotyped using GeneSeek Porcine 50K gene chip;

Fecal samples were collected from the anus of pigs with rectal swabs, and the collected samples were temporarily stored in ice boxes, and then transferred to the laboratory for storage in a -80°C refrigerator.

Further, the data quality control includes microbiome data quality control and genome data quality control;

The quality control of the microbiome data includes: using the DADA2 plug-in in the QIIME2 software to perform quality control and clustering of the original data; performing microbiome data filtering on the data after quality control, and comparing the filtered data to the NCBI RefSeq database. The taxon is annotated for species; the species with a confidence of more than 97% are considered to be the same species, and their relative abundance is calculated;

The filter condition of the microbiome data is: the abundance exceeds 0.1%, and exists in 60% of the samples;

The genomic data quality control includes: using PLINK to filter the original SNP data for genomic data, and SNPs or individuals meeting any of the following conditions will be excluded:

Individuals or SNPs whose deletion rate is greater than 0.1;

SNPs with a minimum allele frequency of less than 0.05;

SNPs that do not fit in Hardy-Weinberg equilibrium.

Further, the correlation analysis includes identification of microorganisms associated with phenotypes and SNP markers associated with microorganisms;

The correlation between the microorganism and the phenotype includes: utilizing the R language to use the CeTF package to realize the partial correlation and the information theory algorithm to calculate the correlation between the relative abundance of the microorganism and the phenotype data;

The identification of the SNP markers related to microorganisms includes: the establishment of a Lasso linear model using the scikit-learn package through python, and the relative relationship between the Terrisporobacter petroleumarius and the Prevotella copri The abundance is the response value, and the SNP data is used as the prediction value, and the SNP sites related to the Terrisporobacter petrolarius and the Prevotella copri are selected.

Further, the verification of the impact of microorganisms and their related SNP sites on the average daily weight gain of pigs includes: using the scikit-learn package to establish different machine learning regression models through python, and performing 10-fold cross-validation respectively, and each fold combines the data It is divided into 30% test set and 70% verification set, and the average daily weight gain is predicted based on the Terrisporobacter petrolearius, the Prevotella copri and their related site information.

The above method can well predict the daily gain of pigs, effectively and accurately select pig breeds with high daily gain, and is conducive to achieving the purpose of improving the production level and economic benefits of the pig industry.

Detailed ways

In order to make the object, technical solution and advantages of the present invention more clear, the present invention will be further described in detail below in conjunction with its embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, and do not limit the protection scope of the present invention.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field of the invention. The terminology used herein in the description of the present invention is only for the purpose of describing specific embodiments, and is not intended to limit the present invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.

In one embodiment of the present invention, a method for predicting daily weight gain of pigs using microorganisms and related SNP sites, the microorganisms include Terrisporobacter petroleumarius and Prevotellacopri.

In one of the embodiments, the SNP sites related to the Terrisporobacter petrolearius include rs339933029, rs333900969, rs332402643, rs338935223, rs80986577 and rs81415286.

In one embodiment, the gene related to the SNP rs332402643 is PDZRN4, the gene related to the SNP rs338935223 is WWOX, and the gene related to the SNP rs80986577 is RAD51B.

In one of the embodiments, the SNP sites related to the Prevotella copri include rs81437804 and rs343769713.

In one embodiment, the gene related to the SNP site rs343769713 is IRS1.

In one embodiment, the method includes the following steps:

Sample collection, microbial DNA extraction and 16S rDNA sequencing, data quality control, correlation analysis and verification of the effects of microorganisms and their associated SNP sites on the average daily gain of pigs.

In one of the embodiments, the sample collection includes the following steps:

When the pigs are 64-150 days old, use the performance automatic measurement system to measure the average daily gain of pigs, and when the body weight reaches 130±5KG, the measurement ends; after the original body weight data is quality controlled, calculate the average daily gain of each individual;

Pig ear tissue was collected for DNA extraction, and the pigs were genotyped using GeneSeek Porcine 50K gene chips;

Fecal samples were collected from the anus of pigs with rectal swabs, and the collected samples were temporarily stored in ice boxes, and then transferred to the laboratory for storage in a -80°C refrigerator.

Preferably, the microbial DNA extraction and 16S rDNA sequencing comprise the following steps:

Microbial genomic DNA extraction: use CTAB to extract the sample genomic DNA;

PCR amplification: pre-denaturation at 98°C for 1 minute; 30 cycles including (98°C, 10 seconds; 50°C, 30 seconds; 72°C, 30 seconds); 72°C, 5 minutes for PCR amplification;

Mixing and purification of PCR products: mix samples at equal concentrations according to the concentration of PCR products, mix well and use 1×TAE concentration of 2% agarose gel electrophoresis to purify the PCR products, select the gel to recover the target band; the product purification kit uses The GeneJET Gel Recovery Kit from Thermo Scientific;

Library construction and on-machine sequencing: Illumina’s TruSeq DNA PCR-Free LibraryPreparation Kit was used for library construction. The constructed library was quantified by Qubit and library detection. After passing the test, NovaSeq 6000 was used for on-machine sequencing.

In one of the embodiments, the data quality control includes microbiome data quality control and genome data quality control;

The quality control of the microbiome data includes: using the DADA2 plug-in in the QIIME2 software to perform quality control and clustering of the original data; performing microbiome data filtering on the data after quality control, and comparing the filtered data to the NCBI RefSeq database. The taxon is annotated for species; the species with a confidence of more than 97% are considered to be the same species, and their relative abundance is calculated;

The filter condition of the microbiome data is: the abundance exceeds 0.1%, and exists in 60% of the samples;

The genomic data quality control includes: using PLINK to filter the original SNP data for genomic data, and SNPs or individuals meeting any of the following conditions will be excluded:

Individuals or SNPs whose deletion rate is greater than 0.1;

SNPs with a minimum allele frequency of less than 0.05;

SNPs that do not fit in Hardy-Weinberg equilibrium.

In one of the embodiments, the correlation analysis includes identification of microbes associated with phenotypes and SNP markers associated with microbes;

The correlation between the microorganism and the phenotype includes: utilizing the R language to use the CeTF package to realize the partial correlation and the information theory algorithm to calculate the correlation between the relative abundance of the microorganism and the phenotype data;

The identification of the SNP markers related to microorganisms includes: the establishment of a Lasso linear model using the scikit-learn package through python, and the relative relationship between the Terrisporobacter petroleumarius and the Prevotella copri The abundance is the response value, and the SNP data is used as the prediction value, and the SNP sites related to the Terrisporobacter petrolarius and the Prevotella copri are selected.

In one of the embodiments, the verification of the influence of microorganisms and their related SNP sites on the average daily gain of pigs includes: using scikit-learn package to establish different machine learning regression models through python, and performing 10-fold cross-validation respectively, each Divide the data into a 30% test set and a 70% validation set, and use the Terrisporobacter petrolearius and the Prevotella copri and their related site information to calculate the average daily weight gain predict.

Preferably, the machine learning regression model includes linear regression LR, random forest RF, support vector machine SVR, gradient boosting tree XGB and decision tree DT.

Embodiments of the present invention will be described in detail below in conjunction with specific examples.

Example 1:

1. Experimental Materials

Taking the Du long three-way hybrid pig as the research object, the growth data, genome data and microbiome data of 385 pigs were collected for use in the present invention. The determination of growth performance was carried out in strict accordance with the internal specifications of the pig farm, and the average daily gain of 385 Du long three-way hybrid pigs was collected.

2. Test method

2.1 Sample collection:

When the pigs were 64-150 days old, the performance automatic measurement system was used to measure the growth traits of the pigs, that is, the average daily gain. When the body weight reached 130±5KG, the measurement was terminated. After quality control of the original body weight data, the average daily weight gain of each individual was calculated.

Pig ear tissues were collected for DNA extraction, and 385 pigs were genotyped using GeneSeek Porcine 50K gene chips.

Fecal samples were collected from the anus of pigs with rectal swabs, and the collected samples were temporarily stored in ice boxes, and then transferred to the laboratory for storage in a -80°C refrigerator.

2.2 Microbial DNA extraction and 16S rDNA sequencing:

(1) Microbial genomic DNA extraction: use CTAB to extract the sample genomic DNA;

(2) PCR amplification: pre-denaturation at 98°C for 1 minute; 30 cycles including (98°C, 10 seconds; 50°C, 30 seconds; 72°C, 30 seconds); 72°C, 5 minutes for PCR amplification;

(3) Mixing and purification of PCR products: Mix the samples at equal concentrations according to the concentration of the PCR products. After thorough mixing, use 1×TAE concentration of 2% agarose gel electrophoresis to purify the PCR products, and select the gel to recover the target band. The product purification kit used the GeneJET Gel Recovery Kit from Thermo Scientific.

(4) Library construction and on-machine sequencing: Illumina’s TruSeq DNA PCR-Free Library Preparation Kit was used for library construction. The constructed library was quantified by Qubit and library detection. After passing the test, NovaSeq 6000 was used for on-machine sequencing .

2.3 Data quality control:

(1) Quality control of microbiome data: First, use the DADA2 plug-in in the QIIME2 (version 2021.4) software to perform quality control and clustering on the original data. Filter the data after quality control (the abundance exceeds 0.1%, and exists in 60% of the samples), and the filtered data is compared with the NCBI RefSeq database to perform species annotation on these taxa; bacteria with a confidence of more than 97% species, can be considered to be the same species of bacteria. After filtering, 18 kinds of microorganisms were obtained, and their relative abundances were calculated.

(2) Genomic data quality control: Use PLINK (version 1.9) to filter the original SNP data, and SNPs or individuals that meet any of the following conditions will be excluded: (1) Individuals or SNPs whose deletion rate is greater than 0.1; ( 2) SNPs with minimum allele frequency (MAF) less than 0.05; (3) SNPs that do not conform to Hardy-Weinberg equilibrium (HWE). After processing, 31,931 high-quality SNP sites were obtained.

2.4 Correlation Analysis

(1) Correlation between microorganisms and phenotypes: We used the R language (version 4.1.3) to implement the partial correlation and information theory (PCIT) algorithm using the CeTF package to calculate the correlation between the relative abundance of the above 18 species and the phenotype data. After calculation, Terrisporobacter petrolearius (Geosporium), Prevotella copri (Prevotella) were significantly correlated with the average daily weight gain.

(2) Identification of SNP markers related to microorganisms: the Lasso linear model was established using the scikit-learn package in python, in which the relative abundance of the two bacteria was used as the response value and the SNP data was used as the predicted value. Selection of SNP sites associated with the two bacteria.

2.5 Verify the influence of microorganisms and their related SNP sites on the average daily gain of pigs:

Use the scikit-learn package to build different machine learning regression models (linear regression LR, random forest RF, support vector machine SVR, gradient boosting tree XGB, decision tree DT) through python, and perform 10-fold cross-validation respectively. Divided into 30% test set and 70% validation set, the average daily weight gain was predicted with the information of Terrisporobacter petrolearius and Prevotella copri and their related loci, and the mean square error ( MSE) reflects the accuracy of prediction, and the smaller the MSE, the more accurate the prediction ability of the model. The MSE of 10 times cross-validation is shown in Table 1 below:

Table 1:

1 time 2 times 3 times 4 times 5 times 6 times 7 times 8 times 9 times 10 times average LR 7.79 8.47 12.03 9.8 14.46 9.9 10.54 9.78 14.61 12.16 10.95 RF 8.67 8.14 15.14 10.78 15.29 8.89 10.72 10.35 15.42 11.55 11.5 SVR 11.82 9.72 12.37 11.58 16.12 11.99 11.72 10.27 19.95 12.49 12.8 XGB 7.64 9.11 12.4 13.82 17.51 10.64 10.81 12.1 15.37 11.8 12.12 DT 16.97 13.1 22.13 15.79 26.31 22.49 15.16 13.78 24.29 20.22 19.02

It can be found from the above table that using the two bacteria Terrisporobacter petrolearius and Prevotella copri and related SNP sites for phenotype prediction, the average predicted MSE ranges from 10.95 to 19.02. etc. Among them, the prediction result of the linear model LR is the best, and the MSE of the decision tree DT is the largest. It can be seen that the daily gain of pigs can be well predicted by using the method of the present invention, and pig breeds with high daily gain can be effectively and accurately selected, which is conducive to achieving the purpose of improving the production level and economic benefits of the pig industry.

The technical features of the above-mentioned embodiments can be combined arbitrarily. To make the description concise, all possible combinations of the technical features in the above-mentioned embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, should be considered as within the scope of this specification.

The above-mentioned embodiments only express several implementation modes of the present invention, and the descriptions thereof are relatively specific and detailed, but should not be construed as limiting the patent scope of the invention. It should be pointed out that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent for the present invention should be based on the appended claims.

Claims (8)

1. A method for predicting daily gain of pigs by using microorganisms and related SNP sites thereof, wherein the microorganisms comprise terrispobacter petrilerielerius and Prevotella (Prevotella copri);

the method comprises the following steps:

sample collection, microorganism DNA extraction and 16S rDNA sequencing, data quality control, correlation analysis and verification of the influence of the microorganism and related SNP loci on the average daily gain of pigs;

the sample collection comprises the following steps:

when the pigs are aged from 64 to 150 days, an automatic performance measuring system is adopted to measure the average daily gain of the pigs, and when the weight reaches 130 +/-5 KG, the measurement is finished; after the original weight data are subjected to quality control, calculating the average daily gain of each individual;

collecting pig ear tissues for DNA extraction, and genotyping pigs by adopting a gene chip of GeneSeek Porcine 50K;

collecting fecal samples from pig anus by using rectal swab, temporarily storing the collected samples in an ice box, and then transferring to a laboratory refrigerator at-80 ℃ for storage.

2. The method of claim 1, wherein the SNP site associated with the Geobacillus terrisporus (Terrisporabacter petrileriarius) comprises rs339933029, rs333900969, rs332402643, rs338935223, rs80986577, and rs81415286.

3. The method of claim 2, wherein the gene associated with the SNP site rs332402643 is PDZRN4, the gene associated with the SNP site rs338935223 is WWOX, and the gene associated with the SNP site rs80986577 is RAD51B.

4. The method of claim 1, wherein the SNP sites associated with Prevotella (Prevotella copri) comprise rs81437804 and rs343769713.

5. The method of claim 4, wherein the gene associated with the SNP site rs343769713 is IRS1.

6. The method of claim 1, wherein the data quality control comprises microbiology data quality control and genome data quality control;

the quality control of the microbiology data comprises the following steps: performing quality control and clustering on the original data by using a DADA2 plug-in QIIME2 software; performing microbiology group data filtration on the quality-controlled data, and performing species annotation on the classification units by comparing the filtered data with an NCBI RefSeq database; the strains with the confidence coefficient exceeding 97 percent are considered as the same strains, and the relative abundance of the strains is calculated;

the microbiology data filtering conditions are as follows: abundance was over 0.1% and was present in 60% of the samples;

the quality control of the genome data comprises the following steps: genomic data filtering of the original SNP data using PLINK, SNPs or individuals that meet any of the following criteria will be excluded:

individuals or SNPs with deletion rates greater than 0.1;

SNPs with a minimum allele frequency of less than 0.05;

not corresponding to SNPs in Hardy-Weinberg equilibrium.

7. The method of claim 1, wherein the correlation analysis comprises identification of phenotypic and microbiologically associated SNP markers;

the phenotype-associated microorganisms include: calculating the correlation between the relative abundance of the microorganisms and the phenotype data by using a CeTF package to realize partial correlation and information theory algorithm by utilizing R language;

the identification of the SNP marker associated with the microorganism includes: the Lasso linear model was established by python using scikit-leann package, and SNP site selection related to the Cladosporium terrestris (terrispobacter petrilerarius) and the Prevotella copri was performed using the relative abundance of the Cladosporium terrestris (terrispobacter petrilerarius) and the Prevotella copri as response values and SNP data as prediction values.

8. The method of claim 1, wherein verifying the effect of a microorganism and its related SNP sites on average daily gain of swine comprises: different machine learning regression models are built by python by utilizing scinit-leann package, 10 folds of cross validation are respectively carried out, each fold of data is divided into 30% of test set and 70% of validation set, and the average daily gain is predicted by the terrispobacter petrilerolericus, the Prevotella copri and relevant site information thereof.