Detection of selective sweep in European wild sheep breeds

Abstract

In wild animal populations, there is a differentiation between populations due to natural selection. The direction and pressure of natural selection in the wild sheep are different in the various geographic areas. Linkage disequilibrium studies showed that regions of the genome in whole wild sheep are under natural selection and that natural selection can affect immune or reproductive or metabolic traits. The study aimed to identify genomic regions under natural selection in wild sheep. For this purpose, the genetic information of 24 European wild sheep and 24 Sardinian wild sheep was used. The genotypes were determined using Illumina 50 K SNPChip arrays based on Oar_4.0 version of the sheep genome. After quality control steps, finally, 31,560 SNP markers were analyzed. The value of LD was calculated by calculating the r² statistic between all pairs of locations through PLINK software. To identify signs of selection based on linkage disequilibrium methods, an extended haplotype homozygosity test of XP-EHH crossing population and iHS intrapopulation was used. The results of iHS studies showed that in European and Sardinian wild sheep, the highest iHS coefficient under natural selection was observed on 3 and 2 chromosome numbers, respectively. Also, the results of XP-EHH studies showed that the largest XP-EHH coefficients under natural selection in European wild sheep compared to Sardinian and vice versa in Sardinian wild sheep compared to European wild sheep were observed on 3 and 16 chromosome numbers, respectively. In addition, the results of gene cycle studies showed that COPB1, SEC24D, ZDHHC17, BBS4, RFX3, SLC26A8, CAMK2D, GRIA1, GRM1, GRID2, PPP2R1A, CPEB4, PLEKHA5 and KIF13A, VPS39, VPS53, DTNBP1, DYNC1I1, FAM91A genes are under natural selection in Sardinian and European wild sheeps, respectively. The direction and selection pressure of natural selection in the two breeds of wild sheep is different due to different geographic conditions.

Introduction

The wild sheep of Asia and Europe, which are the origin of the domestic sheep of the world, are the two wild breeds of Asian mouflon and oriel. The scientific name of the Asian mouflon is Ovis Orientalis and the scientific name of the European mouflon is Ovis Miusimon. Mouflon wild sheep has two populations that live in Europe and Asia. Oriel wild sheep are also divided into two groups, Its eastern group is called Oriel wild sheep and the western group is called Armenian wild sheep (Barbato et al. 2017).

Since wild sheep have adapted to different geographic areas and are therefore resistant to many biotic and abiotic stresses, the genetic diversity in wild sheep is an important genetic source for improving domestic sheep in response to increasing demand for food production, the occurrence of animal diseases, and Global climate change. Therefore, the evolutionary and genetic relationship between wild and domesticated sheep to understand the potential of using wild sheep genetics is important for the improvement of domestic sheep (Chen 2021).

The increase in the frequency of mutations caused by the selection phenomenon, which is useful only in some societies, causes signs to appear at the level of the organism’s genome (Akey 2009). Identifying signs of selection is only possible by examining the allelic frequency, the development of haplotype structure, the reduction of genetic diversity in that area, and the increase of the pattern of genetic linkage disequilibrium (LD) (Qanbari et al. 2012, 2014). Selection is one of the main forces that leads to the creation of marks in certain regions on the genome level. These gene footprints that remain as a result of selection are called selection marks and can be used to identify the positions that have been under selection (Kreitman et al. 2000).

The European mouflon is a wild sheep from the Mediterranean islands of Corsica and Sardinia. It is believed to be a semi-domesticated wild sheep from the Near East that was brought to Sardinia about 8000 years ago. The European mouflon was considered a wild species independent of the European wild sheep. However, recent research has shown that wild sheep have not existed in Europe since the late Pleistocene. Archeologic research in Corsica has shown that the mouflon is not a wild sheep, but a domestic sheep that was brought to the islands by Near Eastern farmers and then turned wild (Grazyna Ptak et al. 2002).

The effective size of the population (Ne) is one of the most important parameters in population genetics due to its direct relationship with the increase in inbreeding and the rate of decrease of diversity in the population. For this purpose, this statistic is considered by a series of research centers such as FAO and IUCN as an important criterion for assessing the risk of livestock extinction (Taberlet et al. 2008).

Moreover, the study of breeds, using molecular techniques is very important and useful for their characterizing (Mohammadi et al. 2009; Mohammadabadi 2021). Conservation of genetic diversity in animal species requires the proper performance of conservation superiorities and sustainable handling plans that should be based on universal information on population structures, including genetic diversity resources among and between breeds (Javanmard et al. 2008; Roudbar et al. 2018). Genetic diversity is an essential element for genetic improvement, preserving populations, evolution, and adapting to variable environmental situations (Mousavizadeh et al. 2009; Masoudzadeh et al. 2020). On the other hand, the determination of gene polymorphism is important in farm animals breeding (Mohammadabadi et al. 2011; Ahsani et al. 2010) define genotypes of animal and their associations with productive, reproductive, and economic traits (Nassiry et al. 2005; Norouzy et al. 2005; Sulimova et al. 2007).

Markers can provide important information regarding the process of evolution and formation of the genome in the target regions of selection (Nielsen et al. 2001, 2005). A new mutation increases the competence of people who carry it compared to other members of society, natural or artificial selection causes people who have more competence to participate more in the formation of the next generation. With the increase in the frequency of beneficial variants, the frequency of alleles in neutral or relatively neutral positions, which are linked to this variant, will also increase (Shabti et al. 2006; Eki 2009). As a result of this phenomenon, the pattern of genetic diversity and linkage LD in the locations around this selective mutant will change. So, the closer to this new allele, the more genetic diversity decreases and LD increases. Therefore, when a beneficial allele becomes the target of positive selection during different times, it creates signs at the genome level that can be identified by examining the spectrum of allelic frequency and LD (Shabti et al. 2006). When a mutation occurs in the genome and has a selective advantage over other mutated loci, to increase the frequency of alleles, it is called positive selection (Kaplan et al. 1989).

When positive selection occurs in a population, the frequency of neutral alleles adjacent to the target allele of selection also increases, and on the other hand, this phenomenon causes an increase in the amount of genetic LD around the selected allele, which leads to the formation of a specific haplotype in the population. (Kim et al. 2002). Therefore, two important phenomena, one is the change of allelic frequency and the other is the imbalance of strong gene linkage between adjacent loci, are two criteria for identifying signs of positive selection (Sabeti et al. 2002, 2006).

One of the methods used to identify areas under positive selection is homozygosity (homozygosity) statistics developed uniformly between populations or XP-EHH. The basis of XP-EHH calculation is by comparing the uniform impurity number developed in two populations for one SNP. It identifies the locations that are homozygous in one population, but polymorphic in another population. In regions under positive selection, an increase in correlation between two or more alleles leads to an increase in the level of linkage disequilibrium (LD) (Nielsen 2005).

The European mouflon is the only non-native representative of the Sahara in the Karkonosze. It was brought here about 100 years ago. It is the ancestor of domestic sheep and the smallest representative of wild sheep. It was introduced in many parts of Europe. The people who were brought to this continent came from Sardinia and Corsica. However, according to some sources, the remaining populations living on those islands, as well as on Cyprus, seem to be descendants of domesticated mouflons that were introduced there in ancient times from their identifying candidate regions under selection helps researchers in understanding the molecular mechanism involved in adaptation and may also be effective and beneficial in identifying regions associated with important traits that are under selection. The purpose of this study was to study genetic diversity and identify the traces of selection in European wild sheep to identify natural selection using single nucleotide markers.

Materials and methods

Data collection

MD SNPs chip (50 K) of 24 European wild sheep and 24 Sardinian wild sheep were considered in this study. The GWAS analysis of these breeds has already been published by Ciani et al 2015. [http://widde.toulose.inra.fr]. There were 48,051 SNP in 27 chromosomes.

Data quality control

To ensure the quality of the data, various stages of quality control were applied to the primary data. Different stages of data quality control were done using PLINK software version 1.09 (Purcell et al. 2007). Also, to have a general view of the population structure, the samples were subjected to principal component analysis (PCA). PCA principal components based on the genomic kinship matrix were performed in the R environment to identify animals that are out of their breed group.

Indicators of genetic diversity

Index of linkage disequilibrium (LD)

LD refers to the non-random association of alleles of different gene loci in populations, which can reflect inherited haplotypes from a common ancestor (Reich et al. 2001). The r² statistic (Hill 1974) was used to estimate LD, which was calculated through the following equation (Hill and Robertson 1968).

$$r^2=\frac{{\left(\text{freq}AB\ast \text{freq}ab-\text{freq}Ab\ast \text{freq}Ba\right)}^2}{\left(\text{freq}A\ast \text{freq}a\ast \text{freq}B\ast \text{freq}b\right)}$$

where freq Ba, freq ab, freq AB and freq Ab show the frequency of haplotypes Ba, ab, AB and Ab, respectively.

The r² statistic was calculated using Plink v1.9 software in R i386 software environment 3.6.2. The r² statistic was evaluated between SNPs at different intervals and the graph of the average r² changes was drawn according to these intervals.

Estimation of effective-population size

Linkage disequilibrium between marker loci was calculated using the criterion of the square coefficient of correlation between two loci r² by the method (Hill and Robertson 1968). The calculated LD r² values were used to estimate the Ne during the past generations.

$$=$$

In this research, EHH statistic was used to check LD in the desired genomic regions. This statistic is a powerful tool that assesses LD erosion around a candidate genomic region by assessing haploid properties within a population (Shabti et al. 2002).

Selection signs

To check the indices of selection markers and to identify the areas under selection in each breed, the wild sheep populations of each breed should be combined based on the similar markers (SNP) between the two populations. Plink v1.9 software was used to combine wild sheep populations.

Within population selection signatures (iHS)

The iHS provides a standard measure of EHH reduction of derived alleles in a region (for example, an SNP) compared to ancestral alleles. Regions with a slow trend of EHH reduction in derived alleles are considered signs of selection.

To investigate the regions carrying selection markers, the integrated haplotype ranking statistic has been used as an intrapopulation statistic. Haplotype files, using the rehh software package (Gautier et al. 2012) in the R software environment, the integrated haplotype ranking values for all markers throughout the genome were calculated using the following relationship.

$$\text{IHS}=\frac{ln\left(\frac{iH{H}_A}{iH{H}_D}\right)-E\left[ln\left(\frac{iH{H}_A}{iH{H}_D}\right)\right]}{\text{SD}\left[ln\left(\frac{iH{H}_A}{iH{H}_D}\right)\right]}$$

Extended haplotype homozygosity (EHH)

After identifying the candidate regions for selection using the iHS statistic, the developed haplotype homozygosity statistic, which is a statistic based on LD and haplotype length, was used to identify the effect of selection in these regions. If these regions have high allelic frequency, haplotype length, and high LD, it indicates positive selection. The value of this statistic was calculated using the following relationship (Sabeti et al. 2007).

$$\text{EHH}=\frac{1}{n_{\mathit{as}}(n_{\mathit{as}}-1)}\sum _{k=0}^{k_{as.t}}{n}_k(n_k-1)$$

where as = SNP central haplotype, N_as = the number of haplotypes carrying the allele, K_as.t = The number of uniquely developed haplotypes carrying alleles from distance marker s to marker t on the same chromosome and N_k = The number of samples carrying the expanded haplotype k. This statistic is based on LD erosion measures haplotype homozygosity around a single nucleotide marker as the center of the haplotype.

Across population selection signatures (XP-EHH)

XP-EHH method, haplotypes in two populations are compared to consider the variation in the amount of recombination in different parts of the genome.

Unlike the two iHS and EHH methods that identify non-stabilized selective signals, it is responsible for identifying the signals that have been stabilized in some sub-populations compared to the whole population. Data that are outside the desired range as an indication-choices are determined (Sabeti et al. 2007).

XP-EHH method was used to investigate the pattern of positive selection at the genome level of these two populations. The most important test based on LD and haplotype length is the XP-EHH method. In this method, it detects genomic regions under selection by alleles with EHH and high frequency (Sabeti et al. 2007). To calculate the XP-EHH statistic, the ancestral status of the different alleles of the used SNPs must be determined. For this purpose, the ancestral status of SNPs was obtained from the Sheep HapMap project http://www.sheephapmap.org. In the XP-EHH method, haplotypes in two populations are compared to consider the variation in the amount of recombination in different parts of the genome.

In this method, genomic regions affected by positive selection are detected by alleles with EHH and high frequency. In the XP-EHH test, haplotypes in two populations are compared with each other, to consider the variation in the recombination rate of different parts of the genome. To identify selection markers in two populations, R × 64 4.0.4 software and rehh package were used in this environment. In this study, we divided the entire genome into 500kbp segments to better identify the regions under selection, and instead of using the XP-EHH value of each SNP, the average XP-EHH value of each segment was used. Markers whose XP-EHH values were in the upper and lower 99th and 95th percentiles were introduced as regions affected by selection.

$$\mathbf{EHH}=\frac{{\sum }_{i=1}^x\left(\begin{array}{c}\mathbf{ni}\\ 2\end{array}\right)}{\left(\begin{array}{c}\mathbf{na}\\ 2\end{array}\right)\left(\begin{array}{c}\mathbf{nA}\\ 2\end{array}\right)}$$

In this equation, na and nA represent the number of haplotypes in which a and A exist respectively, and hx represents the number of separate haplotypes in a genomic region up to the distance x. After estimating EHH, the XP-EHH statistic is obtained from the following equation.

$${\text{XPEHH}}_{\text{scores}}=\text{Ln}\left(\frac{I_A}{I_B}\right)$$

In the above equation, IA and IB are the integral amounts of EHH according to the genetic distance in two populations A and B, respectively. To identify the signs of selection in two populations, R i386 3.6.2 software and rehh package were used in this environment.

Identification of genes

The genomic regions that were in the upper 1% and 5% percentiles of the XP-EHH and iHS coefficients were identified as candidate regions for selection and were further investigated to investigate the existing genes. The information related to the selected regions, including the chromosome number and the position of the selected markers on the chromosome, was used to identify genes that are possibly related to the selected regions.

Gene network

In this step, based on the genes that showed significant differential expression in the previous step, the protein–protein interaction network using ClueGo v2.5.8 to identify significant KEGG pathways and datasets Ovis aries were drawn in Cytoscape software version 1.6.3. This program is an extension of Cytoscape and significantly develops the biologic interpretations of genes (Bindea et al. 2009) and it identifies and draws the gene network based on the correlation value that exists in the expression level of the relevant genes, and identifies the genes that are most related to each other by the common number removed from them.

Results and discussion

Data quality control

After quality control of genomic data, 24 European sheep and 24 Sardinian sheep remained with 30,557 and 39,979 SNPs, respectively. After merging the wild European and Sardinian populations, 31,560 common SNPs remained between the two populations.

Population structures study

The purpose of analyzing the data into main components is to justify the variance in the data. PCA showed that the PC1 component caused the separation of wild sheep species from each other and the PC2 component caused individuals to be placed in their breed. In this study, PC1 and PC2 analysis accounted for 1.79% and 0.38% of the total variance, respectively. According to Fig. 1, it can be concluded that PCA is a suitable analysis for investigating the population structure using genomic data.

Fig. 1.

Principal component analysis (PCA) diagram showing clustering of European sheep in red color and population size 45 and Sardinian sheep in blue color and population size 45. The first and second principal components represent the population classification attributed to the farm origin of each type of sheep. The PCA figure showed that there is a genetic distance between European and Sardinian wild breeds in terms of relationship, which indicates that each wild breed in Europe has a unique genetic structure

The results of PC2 showed that the genetic variation in Sardinian mouflon was more than that of the European one, which is most likely because the sampling done in Sardinian mouflon was different in geographic areas (Fig. 1).

Principal component analysis (PCA) is one of the multivariate methods that are frequently used in population genetic analysis. This method can identify the genetic structure of populations without considering the initial assumption about the genetic model of populations (Jombart et al. 2010; Ringer et al. 2008). However, the evaluation of the genetic groups identified by this method requires an initial definition of the clusters in the population, even though the variance between individuals in the population consists of two components, inter-group variance, and intra-group variance, this method is only focused on the total variance of individuals. However, methods for evaluating the relationship between genetic clusters are more suitable that focus on the variance between groups and ignore the variance within groups (Patterson et al. 2006).

The Sardinian and Corsican mouflon are classified as the European mouflon (Ovis Gmelini musimon) together with mouflon populations introduced from mainland Europe, despite the large genetic distance identified between Sardinian and European mouflon mtDNA lineages. The Corsican–Sardinian mouflon represents an early branch of the evolutionary branch that is the origin of the HPG-B domestic sheep and the original mouflon of mainland Europe. In addition, the Sardinian mouflon gene has the oldest HPG-B haplotype that has been identified so far (Satta et al. 2021).

Since wild sheep have adapted to different geographic ranges and thus are resistant to many biotic and abiotic stresses, the genetic diversity in wild sheep is an important genetic resource for improving domestic sheep in response to increasing demand for food production, the occurrence of animal diseases, and global climate change. Therefore, the evolutionary and genetic relationship between wild and domesticated sheep is important to understand the potential of using wild sheep genetics to improve domesticated sheep (Chen 2021).

Indicators of genetic diversity

Heterozygosity

The heterozygosity index also estimates genetic diversity in a population and is one of the most widely used parameters for calculating genetic diversity in a population. High heterozygosity indicates crossbreeding of non-relatives and low allelic fixation.

The average amount of heterozygosity in the European sheep population was calculated as 0.352. The highest and lowest number of markers for each chromosome were observed for chromosome number 1 and chromosome number 24, equal to 3436 and 403 markers, respectively (Fig. 2).

Fig. 2.

Heterozygosity of SNPs in European and Sardinian sheep. The average amount of heterozygosity in the European sheep population was calculated as 0.352. The highest and lowest number of markers for each chromosome were observed for chromosome number 1 and chromosome number 24, equal to 3436, and 403 markers, respectively (European mouflon). The average amount of heterozygosity in the Sardinian sheep population was calculated as 0.328. The highest and lowest number of markers for each chromosome were observed for chromosome number 1 and chromosome number 24, equal to 4406, and 533 markers, respectively (Sardinian mouflon). The heterozygosity index also estimates genetic diversity in a population and is one of the most widely used parameters for calculating genetic diversity in a population. High heterozygosity indicates crossbreeding of non-relatives and low allelic fixation

The average amount of heterozygosity in the Sardinian sheep population was calculated as 0.328. The highest and lowest number of markers for each chromosome were observed for chromosome number 1 and chromosome number 24, equal to 4406 and 533 markers, respectively (Fig. 2).

Computing the heterozygosity rate for a genotyping data set with a big number of SNPs and a homogeneous sample population can help identify problematic SNPs (Zhao et al. 2018) low heterozygosity may indicate inbreeding and higher heterozygosity may indicate sample contamination.

The number of alleles in a population is one of the main parameters of genetic diversity and the main factor in the production rate of animals because the presence of more alleles in the population provides more possibilities for the combination and rearrangement of genes (Sharifi et al. 2020).

In a study that was conducted to investigate the genetic diversity of five Iranian sheep breeds (Mherban, Mughani, Khorasan Kurdish, Kurdistan Kurdistan, and Sinjabi) using microsatellite markers. The highest diversity was observed in the Mughani race (0.847) and the lowest diversity in the Khorasan Kurdish race (0.744) (Esmaeilkhanian and Banabazi 2006). Also, using microsatellite markers, Moulai et al. investigated the diversity among six Iranian sheep breeds (Lari, Sanjabi, Bakhtiari, Khuzestan Arabian, Qashqai, and Shiraz Kabode). Also, the amount of expected heterozygosity was calculated in some Iranian sheep breeds including Balochi 0.816, Shirvan Kurdish 0.7423, and Khorasan Kurdish 0.7713 (Daneshvar amoli et al. 2018; Naqoyan et al. 2013).

With the progress of science and the expansion of markers and arrays, the study of genetic diversity at the genome level received more attention. In the study that was conducted using snappy markers on five breeds of Australian sheep (Merinos, Hornless Dorset, Leicester, and two mixed breeds), the expected heterozygosity was in the range of 0.30–0.38 and the observed heterozygosity was in the range of 0.30–0.40. 0–30.0 was obtained (Al-Mamon et al. 2015).

The observed and expected heterozygosity estimates in the study conducted on squirrel sheep were equal to 0.9606 and 0.6487, which shows high genetic diversity (Sharifi et al. 2020), in another study, the average observed and expected heterozygosity In squirrel sheep, it was calculated as 0.64 and 0.77, respectively (Rahbar et al. 2016). The difference between the calculated values in the two mentioned studies can be due to genetic drift in sheep, the effect of sample size, or the type of microsatellite markers. be.

Examining the amount of diversity in the Iranian wild species, the results showed that the expected heterozygosity rate is 0.367 and the observed is 0.439, which indicates the appropriate diversity with the limited population of the wild species and inbreeding (Moradi, et al. 2017) the average rare allele frequency in this study was 0.28, which was in agreement with previous studies (Mohammadi et al. 2017; Al Mamoun et al. 2015).

Microsatellite markers are usually used to identify parents and as an excellent option for forming genetic profiles of breeds, calculating the genetic diversity of different populations, and designing and implementing programs to protect endangered populations. It is possible (Sharifi et al. 2020).

Index of linkage disequilibrium (LD)

In genomic selection, LD between known SNP markers and unknown causal mutations is used to estimate genomic modification value (Aliloo et al. 2018), and the genomic prediction equations in a large reference population are used to estimate the genomic correction values for the selected candidates that only have genomic information (Moisen et al. 2016).

Linkage disequilibrium (LD) in wild sheep in this study showed that as the distance between pairs of SNPs increases, the value of LD decreases (Figure). The highest amount of r² was observed at a distance of fewer than 50 kilobases in both breeds of wild sheep, and the greatest decrease of r² was observed at distances of more than 200 kilobases. The higher the level of LD, the more haplotypes inherited from the common ancestor and, as a result, the lower the amount of genetic diversity. According to the shape, the amount of genetic diversity of European wild sheep is higher. Haplotype blocks are regions of the genome where the rate of recombination is low and they represent historical evidence. Nevertheless, it can be concluded that the rate of recombination and selection is lower in Sardinian wild species. To date, many studies have been conducted to calculate LD in sheep populations. In a study conducted on Zandi sheep, the results showed that the average LD in short distances (20–10 kilobytes) was 0.23–0.25 and in long distances (40–60 kilobytes) was 0.16–10. 0/0, which indicates that the LD rate decreases with the increase in the distance between pairs of SNPs (Ghoreishifar et al. 2019).

The efficiency of LD mapping will depend on the amount of LD in the studied population, LD heterogeneity along the genome, marker density, and QTL allelic heterogeneity (Mohammadi et al. 2017). A population with a high LD value requires a lower marker density (Meadows et al. 2008).

In a study conducted on Chinese wool sheep (Merino breed), the average LD in short distances (0–10 kb) was reported to be more than 0.25 and with increasing distance between pairs of SNPs, the amount of LD decreased significantly. (Liu et al. 2017). Also, in another study conducted on Spanish dairy sheep (Chura breed), the results showed a decrease in LD with an increase in the distance between pairs of SNPs (Chitendi et al. 2017).

The difference in our LD value between chromosomes in sheep has been reported by (Gholizadeh et al. 2015; Liu et al. 2017). These differences can be related to different rates of recombination within and between chromosomes, heterozygosity, genetic drift, and the effect of selection of important economic traits (Liu et al. 2017). In addition, within a chromosome, the recombination rate increases from the centromeric region to the telomeric region. This issue can lead to variation in the amount and extent of LD in different regions of the genome (Arias et al. 2009).

Effective population size (Ne)

When the effective size is small, the genetic diversity within the population is also limited, which affects the amount of genetic progress in breeding programs. On the other hand, before taking any action related to the protection of the genetic reserves of a livestock population, it is necessary to obtain information about the effective size and genetic diversity of that breed (Zhao et al. 2014).

To preserve genetic diversity and prevent weakness caused by inbreeding in different populations, the effective size in the short term (5 generations ago) should be at least 100 individuals. Also, in the long term, the effective size should be more than a thousand individuals to maintain the long-term survival of populations, this is known as the 100 per 1000 rule of thumb (Frankham et al. 2014). In this study, the effective-population size for wild sheep species was calculated up to 3300 generations ago (Fig. 3).

Fig. 3.

LD decay and Ne generation ago plots of European and Sardinian mouflon. a The graph shows that the European wild sheep has been under selection more than the Sardinian wild sheep, so the population effective is lower. b Average linkage disequilibrium (LD) decay (r²) from 0 to 300 kb for each of the European and Sardinian sheep breeds included in the analysis. It used the squared correlation coefficient between two loci (r²). LD graph shows that there is a higher LD in the European mouflon population

In research conducted on five breeds of domestic sheep and one breed of wild sheep in Iran using marker information, the Ne was calculated during the previous 4–3500 generations. The results showed that the effective size had a decreasing trend so that 3500 generations ago, it reached the range of 6000–6500 heads to the range of 89 to 9 heads in the previous 4 generations. The highest effective size in the previous four generations was related to the Zell breed (89 heads) and the lowest was related to wild sheep (9 heads) (Moradi et al. 2017). A study was conducted in 2012 to estimate the effective size of the Zel and Lori-Bakhtiari breeds, the results of this research showed that the effective size in both breeds had a decreasing trend, from 4900 heads in the previous 2000 generations to 840 heads in the Zel breed and 534 heads in the Lori-Bakhtiari breed in the previous 20 generations (Moradi et al. 2012).). Also, in another study conducted in 2014 on three breeds of sheep (Sonit, German Merino, and Dorper) using marker information, the results showed a decrease in effective size in three breeds from 2000 to 7 generations ago. So that the effective size of the Sonit breed has increased from 1506 to 207 heads, Merinos breed from 1678 to 74 heads, and the Dorper breed from 1506 to 67 heads (Zhao et al. 2014).

Within population selection signatures

In this research, the integrated haplotype score (iHS) was calculated for all markers in the entire sheep population.

European mouflon

Based on the results of iHS, 93 SNP were selected in European mouflon sheep. These regions are located on chromosomes 1, 2, 3, 4, 6, 7, 9, 12, 13, 15, 18, 19, 20, 22, and 26. The highest iHS coefficient with the lowest P-value was observed in chromosome number 3 (Table 1). (discussed selection signature in European mouflon) (Fig. 4).

Table 1.

Results of within-population selection signatures in European mouflon

CHR	Start	End	nSNPs	Length (Kbp)	iHS	LOGP	Gene
1	52,407,771	85,505,828	9	33,098.06	2.48	1.93	LOC101114579, LRRC8D
2	75,644,823	99,339,620	9	23,694.8	2.1	1.6	PLGRKT, PTPRD, SMARCA2
**3	95,143,985	127,549,955	14	32,405.9	2.39	2.02	BIRC6, EXOC6B, GLIPR1L1, KCNC2, (LOC105614699), PAWR, PPFIA2, PTPRQ
4	9,898,205	15,562,966	6	5664.7	2.44	1.85	CDK14, COL1A2, DYNC1I1, PPP1R9A, SLC25A13,
4	88,773,117	109,338,529	4	20,565.41	2.59	2.17	POU6F2
6	10,086,493	91,138,679	8	81,052.1	2.57	2	KCNIP4, N4BP2
7	12,860,728	67,422,483	5	54,561.7	2.7	2.2	FMN1, GLCE, RASL12, VPS39
9	81,342,298	83,016,880	2	1674.5	2.33	1.72	RIMS2
12	35,049,063	35,941,230	4	892.1	2.63	1.87	EFCAB2
13	32,324,871	34,584,652	3	2259.7	2.41	1.85	ARHGAP12, CUBN
15	11,797,318	15,304,342	4	3507.02	2.59	2.03	MAML2
18	40,403,382	64,950,237	5	24,546.8	2.51	1.94	COCH, DDX24, SETD3
19	11,087,630	18,427,122	6	7339.4	2.42	1.92	ITGA9, MLH1, SRGAP3
20	32,404,117	46,135,173	5	13,731.06	2.49	1.9	CARMIL1, KIF13A
22	46,261,201	49,045,519	3	2784.3	2.48	1.89	ADAM12
26	7,515,168	39,236,791	10	31,721.6	2.74	2	FUT10, LOC114111030

CHR, start, end, nSNPs length, his and LOGP are chromosome number, start coordinate, end coordinate, number of SNPs into the region, length of the region, iHS coefficient and − log10[2Φ − |iHS|], respectively

Fig. 4.

Manhattan plot of genome-wide iHS analyses on European sheep. In Plot of the integrated haplotype score (iHS) plot for European sheep, the horizontal black dashed line marks the significance threshold of − log10 (P value) = 4. If SNPs exceeded this threshold, they were considered as selection signatures

Sardinian mouflon

Based on iHS results, 157 SNP have been selected in European Sardinian sheep. These regions are located on chromosomes 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 15, 16, 18, 19, 25, 26. The highest iHS coefficient with the lowest P-value was observed in chromosome number 2. (discussed selection signature in European mouflon) (Fig. 5) (Table 2).

Fig. 5.

Manhattan plot of genome-wide iHS analyses on Sardinian sheep. In Plot of the integrated haplotype score (iHS) plot for Sardinian sheep. A horizontal black dashed line marks the significance threshold of − log10 (P value) = 4, and if SNPs exceeded this threshold, they were considered selection signatures

Table 2.

Results of within-population selection signatures in Sardinian mouflon

CHR	Start	End	nSNPs	Length (Kbp)	iHS	LOGP	Gene
**2	50,236,460	80,075,988	28	29,839.5	3.3	2.8	GLIS3, LOC101117653, LOC105607911, LOC105607964, PTAR1, RFX3, SMARCA2
3	24,125,200	119,945,460	19	95,820.26	2.85	2.26	CCDC91, DDX31, EIF2AK2, ELMOD3, HEATR5B, NDUFAF7, SULT6B1, ZDHHC17
4	2,036,361	5,885,250	15	3848.8	2.85	2.4	GRM8, LOC101106517, POU6F2
5	67,633,600	94,005,784	12	26,372.1	2.76	2.25	EDIL3, GRIA1
6	36,192,023	72,594,264	10	36,402.24	2.63	2.06	COMMD8, GRID2, PDS5A
7	20,292,100	25,377,175	5	5085.07	2.57	2.05	MYO9A
7	53,993,643	91,589,143	4	37,595.5	2.68	2.13	IFT43, MYZAP
8	63,904,002	70,709,943	3	6805.9	2.61	2.32	STX7
9	1,940,729	7,080,616	9	5139.9	2.93	2.51	ADGRB3, LOC101103487, RIMS1
9	12,334,494	28,340,689	4	16,006.1	2.89	2.44	KHDRBS3
10	16,681,052	17,012,728	3	331.6	2.63	2.06	TSC22D1
11	37,232,614	52,708,895	3	15,476.3	2.55	1.96	BRIP1
12	11,310,194	29,582,074	4	1827.9	2.75	2.24	CAPN2
13	64,639,487	71,300,218	8	6660.7	2.76	1.93	BPIFB2, LOC114117614, LOC114117615, NCOA6, SNTA1
15	26,544,897	79,420,398	9	52,785.5	2.69	2.16	COPB1, IPO7, UVRAG
16	21,245,946	35,866,332	5	14,620.3	2.67	2.11	PDE4D, RAB3C
18	40,275,535	48,319,857	5	8044.3	2.81	2.35	GEMIN2, PAX9
19	5,622,653	5,635,464	3	12.8	3.1	2.7	GADL1
25	4,990,072	23,098,132	5	18,108.06	2.62	2.35	CTNNA3, SIPA1L2
26	4,467,736	5,339,343	3	871.6	2.56	2.2	CSMD1

CHR, start, end, nSNPs length, his and LOGP are chromosome number, start coordinate, end coordinate, number of SNPs into the region, length of the region, iHS coefficient and − log10[2Φ − |iHS|], respectively

Across population selection signatures

The XP-EHH statistic, which is based on LD and haplotype length, was used to identify genomic regions under positive selection. This method detects genomic regions under selection using ancestral alleles and haplotype phases. After the calling stage of SNPs with ancestral information from the total of SNPs, data editing was done. In this research, a Manhattan graph was drawn for better identification of genomic regions under selection in the whole genome (Fig. 6). In this graph, the parts of the genome that have high XP-EHH values indicate the differentiation of populations due to positive selection, which seeks natural and artificial selection during different generations for the desired locations. It has come into existence. The obtained results showed that in several genomic regions, adjacent SNPs had high population differentiation, and in this research, 14 genomic regions were identified on chromosomes 1, 2, 3, 4, 7, 9, 10, 13, 15, 20, and 23, 24 (Table 3). (discussed across population selection signature).

Fig. 6.

Cross-population extended haplotype homozygosity (XP-EHH) plot between European and Sardinian sheep with a significance threshold of − log10 (P). Positive and negative values of XP-EHH coefficients are for European and Sardinian, respectively

Table 3.

Suggested selected SNPs and genes in European and Sardinian sheep based on the XP-EHH method

Breed	CHR	Start	End	nSNPs	Length (Kbp)	XP-EHH	LOGP	Gene
European	1	63,804,532	70,631,985	5	6827.4	2.2	1.97	GBP6
	2	80,343,267	108,775,514	9	28,432.2	2.62	2.06	KDM4C
	3	99,246,368	123,131,503	28	23,885.1	2.58	2.15	ANAPC1, CAPS2, EMX1, EXOC6B, GLIPR1L1, MGAT4A, NAV3, SFXN5, TACR1, TET3, TSGA10, ZC3H6
	4	11,236,528	20,398,963	10	9162.4	2.51	2.09	ANKIB1, BET1, DYNC1I1, PPP1R9A
	7	34,436,508	44,130,188	13	9693.6	2.79	2.15	CAPN3, GANC, L2HGDH, SNAP23, TMEM87A, UBR1, ZNF106
	9	31,416,999	38,409,588	15	6992.5	2.58	2.04	ATAD2, FAM91A1, FBXO32, KLHL38, RB1CC1, XKR4
	11	8,134,841	8,343,359	5	208.5	2.56	2	MFSD11, MGAT5B, MXRA7, ST6GALNAC1
	13	41,314,389	41,443,275	4	128.9	2.66	2.11	RALGAPA2
	15	7,078,006	8,696,202	8	1618.2	2.62	2.05	ANGPTL5, ARHGAP42
	20	18,872,193	23,423,229	5	4551.03	2.6	2.06	MAD2L1BP, MRPS18A
	20	44,735,514	44,834,287	4	98.8	285	2.38	DTNBP1, JARID2
	22	44,615,543	46,263,230	9	1647.7	2.63	2.06	DMBT1, PLEKHA1
	23	39,828,270	39,945,014	4	116.7	2.95	2.49	ROCK1, USP14
	24	18,834,056	19,819,848	3	985.8	2.51	1.92	REXO5
Sardinian	1	256,448,638	258,406,516	3	1957.8	− 2.28	1.75	MBNL1
	2	65,473,406	76,464,705	13	10,991.3	− 2.19	1.8	KCNV2, SMARCA2, TRPM3
	2	208,597,501	229,928,223	6	21,330.7	− 2.28	1.86	CPO, SPAG16
	5	64,831,551	65,381,572	3	550.02	− 2.15	1.56	CCDC69, SLC36A1
	6	13,637,384	14,885,921	12	1248.5	− 2.61	2.2	CAMK2D, UGT8
	7	20,823,806	21,713,359	7	889.5	− 2.2	1.76	ADPGK, BBS4, NEO1
	7	56,880,409	59,878,573	6	2998.2	− 2.16	1.68	MYO5C, UNC13C
	8	56,979,472	77,949,827	7	1834.05	− 2.32	1.45	ADGB, GRM1
	12	29,833,316	29,895,218	3	61.9	− 2.25	1.62	NVL
	13	82,529,766	83,038,788	5	509.02	− 2.2	1.56	ATP9A
	14	61,715,553	61,946,039	3	230.48	− .2.16	1.51	PPP2R1A
	16	5,454,415	7,435,348	24	1980.93	− 2.43	1.89	CPEB4, GFM2, LOC114118722, LOC114118724, NSG2
	17	140,776	1,082,454	7	941.68	− 2.41	1.85	CPE, KLHL2, LOC114108867
	17	3,544,077	5,173,182	9	1629.1	− 2.21	1.75	DCHS2, LOC114108868, TMEM131L

CHR, start, end, nSNPs length, his and LOGP are chromosome number, start coordinate, end coordinate, number of SNPs into the region, length of the region, XP-EHH coefficient and − log10[2Φ − |XP-EHH|], respectively

The standard coat in wild-type sheep is a dark body color with a pale belly (Sponenberg 1997), which is relatively rare in domesticated sheep.

In the ancestors of the wild mouflon (Ovisariesorientalis), due to the lack of artificial selection for the coat color, allows different types to appear and be separated (Norris and Whan 2008) (Table 4).

Table 4.

Gene ontology of identified genes in European and Sardinian sheep

Gene ontology Term	Pathway name	Number of Ggenes in each term	Candidate gene	Adjusted P-value
GO:0005096	GTPase activator activity	8	[ADGRB3, ARHGAP12, ARHGAP42, ELMOD3, MYO9A, RALGAPA2, SIPA1L2, SRGAP3]	0.04
GO:0000149	SNARE binding	5	[ANKRD27, SEC24D, SNAP23, STX7, SYT1]	0.00
KEGG:04730	Long-term depression	5	[GRIA1, GRID2, GRM1, PPP2R1A, RYR1]	0.05
GO:0030695	GTPase regulator activity	8	[ADGRB3, ARHGAP12, ARHGAP42, ELMOD3, MYO9A, RALGAPA2, SIPA1L2, SRGAP3]	0.04
GO:0042383	Sarcolemma	5	[CAPN3, DTNBP1, DYSF, RYR1, SGCG]	0.03
GO:0031674	I band	5	[CAPN3, FBXO32,PPP1R12A, RYR1, SYNPO2]	0.03
GO:0008066	glutamate receptor activity	4	[GRIA1, GRID2, GRM1, RELN]	0.05
GO:0060589	N regulator activity	8	[ADGRB3, ARHGAP12, ARHGAP42, ELMOD3, MYO9A, RALGAPA2, SIPA1L2, SRGAP3]	0.04
GO:0043197	Dendritic spine	4	[DTNBP1, GRIA1, GRID2, PPP1R9A]	0.00
GO:0006906	Vesicle fusion	4	[ANKRD27, DYSF, SYT1, VPS39]	0.04
GO:0016050	Vesicle organization	7	[ANKRD27, DTNBP1, DYSF, KIF13A, SEC24D, SYT1, VPS39]	0.00
GO:0044309	Neuron spine	4	[DTNBP1, GRIA1, GRID2, PPP1R9A]	0.00
GO:0030315	T-tubule	3	[CAPN3, DYSF, RYR1]	0.03
GO:0090174	Organelle membrane fusion	4	[ANKRD27, DYSF, SYT1, VPS39]	0.04
GO:0061025	Membrane fusion	5	[ANKRD27, DYSF, STX7, SYT1, VPS39]	0.04
GO:0006836	Neurotransmitter transport	7	[DTNBP1, RIMS2, SLC25A13, SLC36A1, SNAP23, SYT1, UNC13C]	0.00
GO:0055037	Recycling endosome	5	[ATP9A, DYNC1I1, GRIA1, STX7, VPS53]	0.007
GO:0071470	Cellular response to osmotic stress	3	[CAPN3, DYSF, LRRC8D]	0.03
GO:0032279	Asymmetric synapse	6	[CPEB4, DTNBP1, GRIA1, GRID2, GRM1, PLEKHA5]	0.05
GO:0014069	Postsynaptic density	6	[CPEB4, DTNBP1, GRIA1, GRID2, GRM1, PLEKHA5]	0.05
GO:0007215	Glutamate receptor signaling pathway	4	[GRIA1, GRID2, GRM1, RELN]	0.05
GO:0030133	Transport vesicle	6	[ANKRD27, DTNBP1, GRIA1, RAB3C, SEC24D, SYT1]	0.00
GO:0030018	Z disk	4	[CAPN3, FBXO32, PPP1R12A, SYNPO2]	0.03
GO:0030017	Sarcomere	5	[CAPN3, FBXO32, PPP1R12A, RYR1, SYNPO2]	0.03
GO:0034702	Ion channel complex	7	[GRIA1, GRID2, KCNC2, KCNIP4, KCNV2, LOC101108953, LRRC8D]	0.01
GO:0097447	Dendritic tree	7	[CPEB4, DTNBP1, GRIA1, GRID2, GRM1, PPP1R9A, RELN]	0.05
GO:0030425	Dendrite	7	[CPEB4, DTNBP1, GRIA1, GRID2, GRM1, PPP1R9A, RELN]	0.05

Gene network

Figures 7 and 8 show the gene network under positive selection in Sardinian and European wild sheep breeds. The circles represent the genes and the lines between them are the interactions between them. In this study, it was found that some genes were under selection in Sardinian wild sheep and European wild sheep, and these genes play a role in several important cycles, therefore, these genes are investigated.

Fig. 7.

Result of gene pathway analysis plot of European sheep. Result of gene pathway analysis plot of European sheep. The figure shows the gene network under positive selection in the European sheep wild sheep breed. The circles represent the genes and the lines between them are their interactions

Fig. 8.

Result of gene pathway analysis plot of Sardinian sheep. The figure shows the gene network under positive selection in the Sardinian wild sheep breed. The circles represent the genes and the lines between them are their interactions

Genes under selection in Sardinian wild sheep

Coated protein complex 1 (COPB1) is one of the genes that was determined to be under selection in this study. The role of the COPB1 gene, which is located on chromosome 15 of sheep, has been reported in cycles GO:0030660 and GO:0005798. These gene cycles are related to the activity of the Golgi apparatus in the cell, and according to the results of this research, the COPB1 gene, along with the two genes SEC24D and ZDHHC17, is involved in the mentioned two cycles, it was suggested that the gene related to muscle development in pigs (Qiu et al. 2010) and the processing of membrane proteins (Rohn et al. 2000; Vioti 2016). In research on this gene in Pacific Ocean fish, it was suggested that this gene is probably effective in biologic processes such as cell proliferation, traits related to nutrition metabolisms such as feed efficiency (FER) and feed intake rate (RFI) (Yu et al. 2022). It has also been found in human research that the COPB1 gene is essential in the packaging and transfer of proteins and lipids from the Golgi apparatus to the endoplasmic reticulum, and COPB1 subunits are important in brain development and human health. If this gene is disrupted, microcephaly, cataracts, precerebral abnormalities, and sometimes anophthalmia are observed in mutant animals (Macken et al. 2021). It seems that due to the role of this gene in some nutritional and metabolic traits in the investigated animals, it has been selected.

SEC24D gene located on chromosome 6 of Sardinian wild sheep was selected and it was found that it is involved in two cycles GO:0030660 and GO:0005798. It seems that this gene is related to the activity of the Golgi apparatus in the cell, which affects cell development (Johnson 2022) and the metabolism and processing of membrane proteins (Rohn et al. 2000; Vioti 2016). Some studies in humans have shown that this gene is a strong candidate in connection with cartilage strength and craniofacial congenital defects (Sarmah et al. 2010; Lu et al. 2022). Also, in a study on mice, it was shown that a defect in the SEC24D gene causes early embryonic death (Baines et al. 2013). According to the conducted studies, this gene was probably under selection due to its connection with lamb survival traits in wild sheep.

The ZDHHC17 gene is located on chromosome 3 of sheep, and like the two genes mentioned earlier, it is related to the activity of the Golgi apparatus in the cell in connection with the processing of membrane proteins and the cellular developmental system and metabolism. Studies have shown that the ZDHHC17 gene is expressed in the Golgi membrane, the vesicle membrane associated with the Golgi (ZDHHC17-NCBI), also considering that the expression of this gene is a set of nerve proteins and the lack of expression of this gene causes difficulty in swallowing and animals walk slowly (Yang et al. 2013). Therefore, it seems that this gene has been under selection due to its association with nutritional and digestive traits in wild animals.

The BBS4 gene is located on chromosome 7 of sheep. This gene, together with two genes RFX3 and SLC26A8, plays a role in cycles GO:0060285 and GO:0001539. These two cycles are related to the activity of the reproductive system (Maillo et al. 2016) and spermatogenesis (Kistler et al. 2015). Also, some studies have shown that BBS4 gene disruption is related to obesity and kidney failure (Sheffield et al. 2004; Gorman et al. 1999; Katsanis et al. 2002). It seems that this gene has been under selection due to its relationship with reproductive power and weight control in wild sheep. RFX3 gene on chromosome number 2 in Sardinian wild sheep has been selected, and three single nucleotide SNP mutations have been identified for this gene. This gene is also involved in two cycles GO:0001539 and GO:0060285. Studies have shown that these cycles are related to the activity of the reproductive system in cattle (Maillo et al. 2016) and sperm production (Kistler et al. 2015). This gene has been selected due to its importance in reproduction and related traits. The SLC26A8 gene, which is located on chromosome number 20, like BBS4 and RFX3 genes, plays a role in two cycles 0001539 and GO:0060285. A SNP was identified for this gene. In other reports, the role of this gene in spermatogenesis has been reported. The SLC26A8 domain was identified as a GTPase accelerator in male germ cells (Naud et al. 2003; Toure et al. 2001). It was also observed that the lack of expression of this gene in infertile mice causes a decrease in the amount of ATP and thus immobility of sperm and reduced-fertilization capacity (Toure et al 2007.

CAMK2D gene is located on chromosome 6 and was selected in wild Sardinian sheep. An SNP was observed for this gene. CAMK2D gene along with GRIA1, and GRM1 genes is involved in the KEGG:04720 cycle. According to the role of cycle KEGG:04720, this gene is probably related to the reproductive system (ovulation in female animals). This gene is associated with the long-term potentiation cycle (LTP), LTP is considered a prime candidate for the cellular mechanisms involved in learning and provides an attractive hypothesis of how memories are formed (Kumar 2011), so this gene can have an effective effect on the memory and recognition of the seasonal migration route in wild animals.

The GRIA1 gene is located on chromosome 5. This gene was under selection in Sardinian wild sheep and had an SNP. In addition to its role in the KEGG:04720 cycle, the GRIA1 gene is also important in several other cycles (, KEGG:04730, GO:0032279, GO:0008066, GO:0014069, GO:0007215). These gene cycles are related to long-term depression, asymmetric synapse, and glutamate receptor activity, postsynaptic density, glutamate receptor signaling pathway. This gene, along with GRID2, GRM1, and PPP2R1A genes, plays a role in creating the mentioned cycles. Research shows different effects of genes. Sugimoto et al. 2010 stated that the GRIA1 gene has an effect on GnRH release in addition to LH secretion, so this gene has an effect on reproductive function through ovulation. Although some researchers such as Zamanillo et al. (1999) showed that the GRIA1 gene does not play a role in reproduction in mice, it seems that it is difficult to investigate the role of this gene in animals with multiple ovulations such as mice, and it may play a prominent role in monozygotic animals (Sugimoto et al. 2010).

The GRM1 gene, which is located on chromosome 8, has been under selection in Sardinian wild sheep, and an SNP was observed for this gene. It has been reported that a somatic mutation in this gene causes various types of tumors such as melanoma, breast carcinoma, colon carcinoma, lung adenocarcinoma, and brain, hematopoietic, and lymphatic tissue tumors (Esseltine et al. 2013). Disorders of this gene affect digestion, blood circulation, fertility of male sheep, and lactation of female sheep. Probably, this gene is under selection due to its relationship with reproductive traits and milk production in wild sheep.

The GRID2 gene is located on chromosome 6 and is under selection in Sardinian wild sheep. An SNP was observed in this gene. Different mutations in this gene cause cerebellar ataxia in humans, the patient mainly shows the early onset of cerebellar ataxia, cerebellar atrophy, nystagmus, and developmental delay with the least amount of intellectual disability (Taghdiri et al. 2019) and considering the impact of the mutation This gene has been selected on cerebellum and growth delay, probably due to the effect of this gene on metabolism and growth in wild sheep.

The PPP2RIA gene is located on chromosome 14 of sheep and it has been selected in wild Sardinian sheep. An SNP was observed for this gene. This gene is involved in the cycle KEGG:04730. This cycle affects long-term depression. This gene has been reported to be associated with spontaneous abortion in Holstein cattle (Oliver et al. 2019).

The CPEB4 gene is located on chromosome 16 of sheep and it has been selected in Sardinian wild sheep, and an SNP has been observed for this gene. This gene plays a role in cycles GO:0032279 and GO:0014069, these two cycles are related to asymmetric synapse and postsynaptic density. Considering that this gene regulates mRNA translation (Huang et al. 2006), it was probably chosen because of its role in regulating gene expression.

The PLEKHA5 gene is located on chromosome 5 of sheep and it has been selected in Sardinian wild sheep, and an SNP has been observed for this gene. This gene also plays a role in cycles GO:0032279 and GO:0014069. Liu et al., (2020) reported that PLEKHA5 acts as a tumor suppressor in breast cancer metastasis, this gene is expected to be effective in lactation and proper breast function.

Genes under selection in European wild sheep

The KIF13A gene, which is located on chromosome 20 of sheep, had one SNP. This gene is involved in all related cycles in European wild sheep (GO:0007041, GO:0051648, GO:0051650, GO:0072384, GO:0099518, GO:0047496). These gene cycles are related to lysosomal transport, vesicle localization, the establishment of vesicle localization, organelle transport along the microtubule, vesicle cytoskeletal trafficking, and vesicle transport along the microtubule. In the mentioned cycles, this gene cooperates with DTNBP1, DYNC1I1, FAM91A1, KIF13A, VPS39, and VPS53 genes. It has been reported that the KIF13A gene, as a member of the kinesin-3 family, plays an important role in the production of RE tubules by interacting with several proteins (Shakya et al. 2018; Delevoye et al. 2014).

VPS39 gene is located on chromosome 7 of sheep, this gene had one SNP. It plays a role in the cycle GO:0007041, this gene cycle is related to lysosomal transport. It has been reported that this gene can be effective in the animal’s movement and gait, the lack of VPS39 contributes to impaired muscle differentiation and reduced glucose absorption (Davegardh et al. 2021).

VPS53 gene is located on chromosome 11 of sheep, this gene had one SNP. This gene plays a role in the GO:0007041 cycle, which is related to lysosomal transport. Functional mutations in this gene may contribute to neurodevelopmental disorders, and therefore this gene is probably of critical importance in cellular and organismal physiology (Shi et al. 2019).

DTNBP1 gene is located on chromosome 20 of sheep, this gene had three SNPs. It plays a role in the cycles GO:0051650, GO:0072384, GO:0099518, GO:0047496, and GO:0051648, this gene cycle is related to vesicle localization, the establishment of vesicle localization, organelle transport along the microtubule, vesicle cytoskeletal trafficking and vesicle transport along is microtubule. It has been reported that this gene is associated with unrestrained behavior in wild animals, as this gene is associated with schizophrenia (Riley et al. 2009).

The DYNCIII gene is located on chromosome 4 of sheep, this gene had an SNP and is involved in the same cycles as that expressed for the DTNBP1 gene. Considering the role of this gene in egg maturation (Davegardh 2008), it was probably selected due to the role of this gene in reproductive traits in European wild sheep.

FAM91A1 gene is located on chromosome 9 of sheep, this gene had one SNP. It is involved in the cycle of GO:0051648 and GO:0051650. This gene has been reported to be involved in intracellular protein transport and vesicle binding to the golgi. Also, the expression of this gene is related to the cell metabolism cycle in sheep (https://www.ncbi.nlm.nih.gov/gtr/genes/157769/).

Conclusion

The results showed Heterozygosity of SNPs in European and Sardinian sheep had a downward trend. The direction of selection caused the stabilization of alleles related to higher production and the targeting of mating to achieve higher production. This reduction of heterozygosity will stabilize the allele by increasing the number of alleles in homozygous individuals. In this study, the coefficient of correlation and consanguinity increased for several generations of in-group selection, this caused a greater percentage of homozygous gene loci and reduced genetic diversity. The results of LD studies are the first step in determining the appropriate marker density and sample size in genomic studies. Therefore, the preparation of an LD map is inevitable to increase accuracy. The results of iHS studies showed that in European and Sardinian wild sheep, the highest iHS coefficient under natural selection was observed on 3 and 2 chromosome numbers, respectively. Also, the results of XP-EHH studies showed that the largest XP-EHH coefficients under natural selection in European wild sheep compared to Sardinian and vice versa in Sardinian wild sheep compared to European wild sheep were observed on 3 and 16 chromosome numbers, respectively. In addition, the results of gene cycle studies showed that COPB1, SEC24D, ZDHHC17, BBS4, RFX3, SLC26A8, CAMK2D, GRIA1, GRM1, GRID2, PPP2R1A, CPEB4, PLEKHA5 genes were under natural selection in Sardinian sheep and genes KIF13A, VPS39, VPS53, DTNBP1, DYNC1I1, FAM91A1 were under natural selection in wild European sheep. the analyses show possible Selective Sweep in pressures, which are mainly associated with the effects of environmental factors, on the immune response. The direction and selection pressure of natural selection in the two breeds of wild sheep is different due to different geographic conditions. Linkage disequilibrium showed that regions of the genome in any wild sheep are under natural selection and that natural selection can affect immune reproductive or metabolic traits.

Funding

This work was supported by the University of Torbat Heydarieh [Grant Numbers 141, 09/07/2022].

Data availability

The data supporting the findings of this study are presented in various tables and figures and is available within the article.

Declarations

Conflict of interest

The authors declare that they have no conflict of interest in the publication.

Ethical approval

No Animal Care Committee approval was necessary for the purposes of this study, as all information required was obtained from pre-existing databases. (http://widde.toulouse.inra.fr).