Genome-Wide Detection of Copy Number Variations Associated with Miniature Features in Horses

Copy number variations (CNVs) are crucial structural genomic variants affecting complex traits in humans and livestock animals. The current study was designed to conduct a comprehensive comparative copy number variation analysis among three breeds, Debao (DB), Baise (BS), and Warmblood (WB), with a specific focus on identifying genomic regions associated with miniature features in horses. Using whole-genome next-generation resequencing data, we identified 18,974 CNVs across 31 autosomes. Among the breeds, we found 4279 breed-specific CNV regions (CNVRs). Baise, Debao, and Warmblood displayed 2978, 986, and 895 distinct CNVRs, respectively, with 202 CNVRs shared across all three breeds. After removing duplicates, we obtained 1545 CNVRs from 26 horse genomes. Functional annotation reveals enrichment in biological functions, including antigen processing, cell metabolism, olfactory conduction, and nervous system development. Debao horses have 970 genes overlapping with CNVRs, possibly causing their small size and mountainous adaptations. We also found that the genes GHR, SOX9, and SOX11 may be responsible for the miniature features of the Debao horse by analyzing their overlapping CNVRs. Overall, this study offers valuable insights into the widespread presence of CNVs in the horse genome. The findings contribute to mapping horse CNVs and advance research on unique miniature traits observed in the Debao horse.


Introduction
Copy number variations (CNVs) are a type of change in the structure of the genome in which segments of DNA sequence variants spanning from 50 bp to several Mb are widely distributed throughout the genome.Compared to the reference genome, CNVs have largescale insertions, deletions, duplications, inversions, and translocations [1][2][3][4].CNVs are common in both human and animal genomes.CNVs are a significant source of genomic diversity and have been related to phenotypic variation, environmental adaptability, and disease susceptibility [5][6][7][8] in humans and livestock [9].CNVs play a role in the adaptation of horse populations to different environments, distinctive phenotype features, population diversity, and evolution.Specific CNVs influence susceptibility or resistance to certain equine diseases by affecting the function or expression of associated genes.CNVs also impact the genes that are crucial for the interaction between the horse and its microbiome, influencing digestion, nutrient absorption, or susceptibility to infections [3,[10][11][12][13].Previous genome-wide association investigations have found a significant relationship between copy-number variants and the development of short stature in horses [14][15][16][17].Despite this, much remains to be learned about the comprehensive copy number differences in the genome-wide mapping of horses, specifically focusing on identifying genomic regions associated with miniature features in horses.To reach this goal, data from whole-genome resequencing was used to make a horse genome-wide CNV map.
Over 45-55 million years ago, horses (Equus caballus), a member of the Equidae family, developed from a small multi-toed mammal to a substantial single-toed mammal.It is a biological model for studying exercise physiology, adaptive responses, and human disorders like infertility and inflammatory diseases [40,41].China has the world's largest population of horses, primarily found in rural areas, but their numbers are rapidly decreasing.Horses have enormously impacted China's political, economic, and military history [42].The Debao horse is a Chinese horse breed with a miniature body size.Debao horses have unusual morphoanatomical adaptations to aid in hilly activities [43,44].The Baise horse, which is also a mountain breed from China's Guangxi Zhuang Autonomous Region, is known for its muscular body, strong legs, and unique head profile, making it versatile in transportation, agriculture, and riding [3].On the other hand, Warmblood horses, known for their calm temperament and agility, excel in equestrian sports like dressage, show jumping, and eventing.Warmblood horses were introduced to China from Germany [42].
Systematic human selective breeding of horses has emphasized conformational and performance traits, creating highly specialized breeds that are frequently identifiable by the highest possible height at the withers [45,46].Horses exhibit significant body size and wither height diversity due to extensive breeding and artificially selected breeding throughout domestication.Body size is currently one of the essential considerations in evaluating and classifying various breeds and a necessary component in breeding initiatives to improve marketability and performance [17,44,47].Body size is also one of the most significant quantitative traits under evolutionary investigation and varies significantly between breeds and populations within species [34].Presently, molecular components associated with body size have been studied in farm animals and humans.Human body height was found to be associated with two chromosomal regions around the genes LCORL/NCAPG and ZFAT [14].Previous research has identified a small number of loci for the genetic basis of size variation in Western ponies and horses [12,34,[48][49][50].The genes MRas and ATF3 in the GHR shed light on the genetics of the Debao pony of small stature.To understand the genetic basis of short size, genes such as ECM, ALPL, IHH, MMP23, TIMP4, COL10A1, Sox9, Sox6, Sox8, Sox11, COL2A1, and COL9A1 have been identified in the juvenile stage of the Debao horse [51].A recent study showed that the genes TBX3, TBX5, and HMGA2 play a role in the genetic underpinnings of Chinese pony dwarfism [52].
The purpose of this study was to use whole-genome resequencing data to establish horse genome-wide CNV mapping.Furthermore, we concentrated on the Debao horse population to identify candidate genes associated with miniature features.Our study advances the Debao horse's breeding and conservation efforts by presenting new data on the genome-wide CNVs of horses.

Sample and Sequencing Information
The horse breeds studied were Debao (DB) horses, Baise (BS) horses, and Warmblood horses from the Animal Husbandry Research Institute in Guangxi Zhuang Autonomous Region, China.Twenty-six blood samples, ten samples from Debao, ten samples from Baise, and six samples from Warmblood horses were extracted using the DNeasy Blood & Tissue Kit (Qiagen, Beijing, China), and DNA integrity was assessed using aga-rose gels for whole-genome resequencing.PE sequencing libraries were created and then sequenced on the Illumina NovaSeq 6000 platform, producing 150-bp reads with an in-tended depth of 10× coverage.All resequencing data were from our previous study [53].

Data Processing
The data processing pipeline was detailed in our previous study [53].Briefly, the FastQC tool v0.12.1 (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/, viewed on 4 April 2023) with the default settings were used to assess the quality of our data.We applied BWA [54] with the parameters 'mem 4 -k 32 -M' to map each sample's clean reads against our newly reconstructed horse genomic data.SAMtools v1.18 was used to convert alignment files to BAM files [55].Picard3.0.(http://broadinstitute.github.io/picard/,accessed on 4 April 2023) is also utilized to filter and eliminate duplicates from the aligned BAM data.The resulting sequencing quality was sufficient for further analysis.

Screening of the Initial Set of CNVs
Genome-wide copy number variation analysis on horse next-generation sequencing data was carried out using CNVcaller [56].To extract CNVs, the data were separated into three groups based on the three horses: Debao (DB), Baise (BS), and Warmblood (WB).Since the quality of the reference genome significantly influences the detection of CNV, in this study, we used the assembled Debao horse genome from our previous study [53] as reference genomes.It divided the reference genome according to the window size of 800 bp to construct the reference genome database.The number of reads in the window and the reads on the alignment corrected the copy number, corrected and standardized the GC content, and finally, the genotype was determined.All other statistical analysis was calculated using the R 4.2.2 package.

Merge CNVR Collections
Based on the screened CNV data information, the boundaries of CNVRs are initially judged according to the distribution of absolute copy numbers, the frequency of mutations, and the significant correlation of adjacent windows, and finally, the adjacent CNVRs with significant correlations in the distribution of copy numbers in the population were merged using the merge function in BEDTools v2.25 [57] to obtain the final copy number variation detection result.In addition, this study also filters the CNVR data set according to the Silhouette score and chromosome.The silhouette score was calculated based on the silhouette coefficient using the R 4.2.2 package.The silhouette coefficient is an evaluation method for the clustering effect.In this study, according to the silhouette coefficient greater than 0.6, the CNVR set was screened, and the CNVR located in the sex chromosome and Scaffold region was removed.

CNVR Annotation and Functional Enrichment Analysis
Based on our reference genome of the Debao horse, the NCBI Genome Database was used to obtain the gene content of the identified CNVRs.The Ensemble database was utilized to identify the genes involved within or partly overlapping with the horse CNVRs using ANNOVAR software [58].Functional enrichment analysis of the CNVRharboring genes was performed using human orthologs to better understand these CNVRs' functions.To determine their functional enrichment, we used the DAVID program (version 2023q1) [59] to conduct Gene Ontology (GO) categorization [60] and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway annotation [61] on these CNVs genes.

Genome-Wide Identification of CNV in Three Horse Breeds
Based on our previous study [53], ten Debao horses, ten Baise horses, and six Warmbloods resequencing data showed an average depth of around 10.87× (Table S1), which were used to detect CNV in horses at the genome-wide level.After filtering, 26 samples of three horse breeds generated 18,974 CNVs; the figures in parenthesis represent the CNV counts of a particular horse breed (Table 1).The DB horses had the most (6564) CNVs, including 4028 breed-specific CNVs, whereas the WB horses had the lowest number.CNVs ranged in length from 1599 to 891,199 bp, with an average length of 5364.85 bp per breed.The maximum length of CNV in DB horses was 891,199 bp, and the average CNV in WB horses was 5598.86 bp.The Warmblood horse population has the highest average number of individual CNVs detected (984.83) and unique breed-specific CNVs (613.33).

Diversity Analysis of CNVRs in Different Breeds
After assembling overlapping CNVs, 4279 CNVRs were identified in 31 pairs of autosomes from the three horse breeds (Figure 1).Among them, the DB horse population shows the largest number of CNVRs (1480), while the WB breed presents the least (1350) (Table 2) due to the intensely artificial selection.The figures in brackets represent breedspecific CNVR counts.We defined a breed-specific CNVR as any CNVR found exclusively in one breed and not shared by any other breed.Genome-wide CNVRs from the three populations were summarized in this study.BS, DB, and WB had 978, 986, and 895 unique CNVRs, respectively, whereas 202 CNVRs were found in all three horse breeds (Figure 2).
tribution throughout horse breeds.Breed-specific CNVs are shown in blue.Green denotes CNV shared by three breeds, whereas red denotes CNVs shared by two breeds.Genome-wide CNVRs from the three populations were summarized in this study BS, DB, and WB had 978, 986, and 895 unique CNVRs, respectively, whereas 202 CNVR were found in all three horse breeds (Figure 2).

Merge and Functional Analysis of CNVRs from Different Horse Breeds
Since different breeds of horses have the same CNVR, a complete map of the CNVR and a comparative overview of their genomes were created by removing the repeated CNV regions between breeds (Table 3 and Figure 3).After the deletion of duplicat CNVRs, 1545 CNVRs were generated from the genomes of 26 horses.The lengths of th identified CNVRs varied significantly, with an average length of 11,178.55bp.Th number of CNVRs on chromosome 20 is the highest among the 31 autosomes, wherea the number on chromosome 31 is the lowest (Figure 4).The CNVR scales on each chro mosome were significantly varied.Chromosomes 6, 7, 12, 13, 20, 22, 23, 28, and 29 con tained dense CNVRs that covered more than 1% of genomic sequences, with chromo some 12's CNVRs encompassing 2.62% of the chromosome's sequences, while chromo some 31 covered only 0.23% (Table 3).

Merge and Functional Analysis of CNVRs from Different Horse Breeds
Since different breeds of horses have the same CNVR, a complete map of the CNVR and a comparative overview of their genomes were created by removing the repeated CNV regions between breeds (Table 3 and Figure 3).After the deletion of duplicate CNVRs, 1545 CNVRs were generated from the genomes of 26 horses.The lengths of the identified CNVRs varied significantly, with an average length of 11,178.55bp.The number of CNVRs on chromosome 20 is the highest among the 31 autosomes, whereas the number on chromosome 31 is the lowest (Figure 4).The CNVR scales on each chromosome were significantly varied.Chromosomes 6, 7, 12, 13, 20, 22, 23, 28, and 29 contained dense CNVRs that covered more than 1% of genomic sequences, with chromosome 12's CNVRs encompassing 2.62% of the chromosome's sequences, while chromosome 31 covered only 0.23% (Table 3).There was no correlation between the incidence rate of CNVRs and chromosomal length (Figure 4) since the longest chromosome 1 had only 89 CNVRs, while chromosome 20 (approximately one-third the length of chromosome 1) had 105 CNVRs.The findings of this study coincide with those of prior studies [13].
Statistics of CNVRs on equine autosomes.The length of CNVRs has been classified into six regions: 1-2 Kb, 2-3 Kb, 3-5 Kb, 5-8 Kb, 8-12 Kb, and >12 Kb (Figure 5).The greater the magnitude of the copy number variation (CNV), the higher the level of ease in its detection.

Functional Annotation of CNVR in Horse
We performed functional analysis and grouped these CNV-affected genes to determine how CNVs might affect gene biotypes in horses.The functional annotation of 1545 CNVRs from 26 samples showed that 72.23% of CNVRs were in the intergenic region of the genome, 22.89% were in the intron region of the genome, and 1.90% were in the exon region of the genome (Figure 6).Molecular function is mainly related to neurotransmitter receptor activity, calcium ion binding, amino acid activity, protein kinase activity, and the heme binding site.Regarding cellular components, these genes were highly enriched in the cytoplasm, plasma membrane, synapses, and cell surface.Pathway analyses in the Kyoto Encyclopaedia of Genes and Genomes (KEGG) showed that it annotated many genes with known highly variable copy numbers that significantly enriched antigen processing and presentation, functions such as cell metabolism, olfactory conduction, and nervous system development (the corrected p-value is less than 0.05 and the number of genes is greater than or equal to 10) (Table 4).There was no correlation between the incidence rate of CNVRs and chromosomal length (Figure 4) since the longest chromosome 1 had only 89 CNVRs, while chromosome   There was no correlation between the incidence rate of CNVRs and chromosomal length (Figure 4) since the longest chromosome 1 had only 89 CNVRs, while chromosome  20 (approximately one-third the length of chromosome 1) had 105 CNVRs.The findings of this study coincide with those of prior studies [13].
Statistics of CNVRs on equine autosomes.The length of CNVRs has been classified into six regions: 1-2 Kb, 2-3 Kb, 3-5 Kb, 5-8 Kb, 8-12 Kb, and >12 Kb (Figure 5).The greater the magnitude of the copy number variation (CNV), the higher the level of ease in its detection.

Functional Annotation of CNVR in Horse
We performed functional analysis and grouped these CNV-affected genes to determine how CNVs might affect gene biotypes in horses.The functional annotation of 1545 CNVRs from 26 samples showed that 72.23% of CNVRs were in the intergenic region of the genome, 22.89% were in the intron region of the genome, and 1.90% were in the exon region of the genome (Figure 6).Molecular function is mainly related to neurotransmitter receptor activity, calcium ion binding, amino acid activity, protein kinase activity, and the heme binding site.Regarding cellular components, these genes were highly enriched in the cytoplasm, plasma membrane, synapses, and cell surface.Pathway analyses in the Kyoto Encyclopaedia of Genes and Genomes (KEGG) showed that it annotated many genes with known highly variable copy numbers that significantly enriched antigen processing and presentation, functions such as cell metabolism, olfactory conduction, and nervous system development (the corrected p-value is less than 0.05 and the number of genes is greater than or equal to 10) (Table 4).Similarly, We used Gene Ontology (GO) analysis to assess the possible biological implications of these CNV genes.GO enrichment was shown to be significant in three areas: biological process, cellular component, and molecular function (the corrected p-value less than 0.05, number of genes more than or equal to 10) (Table 5).Antigen processing and presentation, chemical synaptic transmission, cell morphogenesis, negative regulation of apoptotic processes, intracellular message transmission, and homophilic cell adhesion via plasma membrane adhesion molecules are representative biological processes.Previous studies supported our results [3,5].
The pathways are illustrated with a bubble diagram.The names of the pathways are given in the legend on the left.The abscissa displays the enrichment factor, which is the ratio of the percentage of genes identified to a pathway in a particular gene to the overall balance of genes designated to that route.The higher the enrichment factor, the more the differentially expressed genes in this pathway are enriched.The size and color of the dots denote the number of enriched genes and the level of significance, respectively (Figure 7).Similarly, We used Gene Ontology (GO) analysis to assess the possible biological implications of these CNV genes.GO enrichment was shown to be significant in three areas: biological process, cellular component, and molecular function (the corrected p-value less than 0.05, number of genes more than or equal to 10) (Table 5).Antigen processing and presentation, chemical synaptic transmission, cell morphogenesis, negative regulation of apoptotic processes, intracellular message transmission, and homophilic cell adhesion via plasma membrane adhesion molecules are representative biological processes.Previous studies supported our results [3,5].In this study, 1089 genes (excluding LOC genes) were annotated in the CNVR data set in 26 samples of three horse breeds (Tables S2 and S3), and 970 genes (excluding LOC genes) were annotated in the CNVR data set in 10 Debao horse samples (Tables S4 and S5).Many genes with known highly variable copy numbers were annotated based on the EquCab 3.0 genome.These genes were significantly enriched in antigen processing and presentation, cell metabolism, olfactory conduction, nervous system development, and other functions (the corrected p value is less than 0.05, and the number of genes is greater than or equal to 10).
given in the legend on the left.The abscissa displays the enrichment factor, which is th ratio of the percentage of genes identified to a pathway in a particular gene to the overa balance of genes designated to that route.The higher the enrichment factor, the more th differentially expressed genes in this pathway are enriched.The size and color of the do denote the number of enriched genes and the level of significance, respectively ( In this study, 1089 genes (excluding LOC genes) were annotated in the CNVR da set in 26 samples of three horse breeds (Tables S2 and S3), and 970 genes (excluding LO genes) were annotated in the CNVR data set in 10 Debao horse samples (Tables S4 an  S5).Many genes with known highly variable copy numbers were annotated based on th EquCab 3.0 genome.These genes were significantly enriched in antigen processing an presentation, cell metabolism, olfactory conduction, nervous system development, an other functions (the corrected p value is less than 0.05, and the number of genes is great than or equal to 10).
Through GO analysis of these CNV genes, we found that the following gene (CYP2C108, PLA2G4A, ITGA8, GHR, CYP11B1, CYP2B46, BCL2A1, RBFOX, PLAAT CAPN9, SULF2, CAPN2, ADAMTS15, NOX3, and PCDH18) were associated with th body structure in all horse breeds.Additionally, a distinct set of genes (GRID2, SOX CYP2B6, FBP1, FBP2, MAPKAP1, STX17, TBCA, and SOX11) were found to be associate with body structure in the Debao horse breed.We analyzed our data using KEGG Path way Database annotations.We found that eleven genes were associated with the hors body structure: six of them (CYP2C108, PLA2G4A, ITGA8, GRID2, CYP11B1, an CYP2B46) were related to the three breeds here analyzed, while five genes (GHR, SOX CYP2B6, FBP1, and FBP2) only to the Debao horse.Based on gene ontology and genet pathway research, we hypothesize that the GHR and ADAMTS15 genes in CNVs hav Through GO analysis of these CNV genes, we found that the following genes (CYP2C108, PLA2G4A, ITGA8, GHR, CYP11B1, CYP2B46, BCL2A1, RBFOX, PLAAT1, CAPN9, SULF2, CAPN2, ADAMTS15, NOX3, and PCDH18) were associated with the body structure in all horse breeds.Additionally, a distinct set of genes (GRID2, SOX9, CYP2B6, FBP1, FBP2, MAPKAP1, STX17, TBCA, and SOX11) were found to be associated with body structure in the Debao horse breed.We analyzed our data using KEGG Pathway Database annotations.We found that eleven genes were associated with the horse body structure: six of them (CYP2C108, PLA2G4A, ITGA8, GRID2, CYP11B1, and CYP2B46) were related to the three breeds here analyzed, while five genes (GHR, SOX9, CYP2B6, FBP1, and FBP2) only to the Debao horse.Based on gene ontology and genetic pathway research, we hypothesize that the GHR and ADAMTS15 genes in CNVs have some relationship with miniature features in horses.GHR, SOX9, and SOX11 genes are shown to overlap with CNVRs and play essential regulatory functions in the pathway.

Discussion
A significant amount of study has been undertaken in recent years on the genetic variants underlying the genetic diversity of horse breeds.Furthermore, the prevalence of CNVs in horses and their effect on phenotypic variation remain understudied.In the current study, we studied copy number variations among horse breeds from China using data from whole-genome resequencing.The recent investigation revealed more CNVs than previous studies [3,13,17,34,62].This could be due to different experimental sample sizes or CNV detection technologies.This study's findings on CNV length ranges align with previous research [3].We found that each horse shared a significant percentage of its CNVs with at least one other horse, which is consistent with their results.Our results provide relevant insights that complement the study of [13].In this study, we identified 202 unique CNVs across all horse breeds, all of which were breed-specific.Additionally, we observed the presence of CNVs ranging from 253 to 986, which two or more horse breeds shared.The recent study not only covered a significant number of samples, but it also identified CNV fragments comprising a higher percentage of the genome, resulting in more CNV variation as well as understanding than earlier studies on CNVs in horse breeds in China.This may be due to the diverse research platforms, experimental sample sizes, and detection methodologies of CNVRs [3,5,43].
The distribution of CNVRs in this study is consistent with that discovered in other studies.Chromosome 20 had the most significant number of CNVRs, while Chromosome 12 had the highest proportion of its genome covered by CNVRs.According to an earlier study, the percentage of CNVRs is higher in shorter lengths, while the ratio is lowest in long-length CNVRs [3,13,43,62,63].These findings suggest that the proportion of CNVR differs across breeds, likely attributable to the genetic origins of these populations or the varied samples gathered for each breed.This observation may be addressed by the higher frequency of shorter CNVRs in the equine genome [3,64,65].
Many exhaustive CNV studies regarding horses have been undertaken, focusing mainly on CNV discovery and reporting their numbers and categories.In contrast, just a few studies looked at the possible relationship between CNVs and complex features.Several studies have shown that genes that overlap with CNVs influence the characteristics of domestic animals.For example, some coat colors in horses, pigs, and sheep are controlled by genes that are impacted by CNVs [3,66].Further research has revealed that CNV-affected genes also contribute to adaptations in horses [3,13], and some are found to influence growth and body size in horses [17,33,49].Studies on cattle have shown a relationship between growth features such as cannon circumference, body slanting length, chest girth, and body weight and the loss and normal CNV types of the CLCN2 gene [67].Copy number variant regions (CNVRs) were explored in order to better understand the genes linked with adaptability (BLA-DQB, EGLN2) and coat color (KIT, MITF) in cattle [2].
Surprisingly, some CNVRs were regarded as breed-specific or were identified in a few species, indicating that the variability may contribute to determining phenotypic characteristics and differentiating between variety-specific traits [68].In the present study, breed-specific CNVRs were identified, and the most significant number of breed-specific CNVRs was detected in the Debao horse.GHR expression was comparatively low in the long bones of the Debao horse.By downregulating the MAPK signaling cascade, GHR may reduce Debao horse height.The candidate genes Sox9 and Sox11 were found to promote the early occurrence of bio mineralization and epiphyseal closure in Debao ponies [51].The candidate gene ADAMTS15 is associated with muscle development, such as myoblast fusion in thoroughbred horses [69].ADAMTS have been reported to play an essential role in damaged cartilage tissues and the musculoskeletal system and restrict locomotor functions in horses [70].
Based on the overlap of CNVRs, we found that the genes GHR, SOX9, and SOX11 may have miniature features in the Debao horse.We also assume that the GHR and ADAMTS15 genes in CNVs are related to small body size-related traits in the studied horse breed.They were also screened out based on biofunction, and they may have been identified as candidate genes for the miniature features in the Debao horse.

Conclusions
In this study, whole-genome resequencing data at the level of the genomic structure were used to make a complete and detailed CNV-based genome-wide model map of horses.Numerous distinct and breed-specific CNVRs in horse breeds support the idea that CNVs are common in the horse genome.Our analysis of the overlapping CNVRs of genes GHR, SOX9, and SOX11 has revealed that these may be responsible for the Debao horse's miniature features.Furthermore, research is necessary to determine the roles of these types of genes in horses.These findings contribute to our understanding of CNVs in the horse genome and have potential implications for future research, providing new insights into the molecular mechanisms underlying miniature features in horses.
Tables S4 and S5: KEGG pathway analysis and GO pathway analysis in the Debao horse are available as Supplementary Materials.Institutional Review Board Statement: The Agricultural Genomics Institute at Shenzhen, Chinese Academy of Agricultural Sciences (AGIS-CAAS) and the Guangxi Veterinary Research Institute obtained the samples following the guidelines approved by the animal ethics committee.
Informed Consent Statement: Not applicable.

Figure 1 .
Figure 1.The distribution of CNVs in various horse breeds.Breed-specific and shared CNV distribution throughout horse breeds.Breed-specific CNVs are shown in blue.Green denotes CNVs shared by three breeds, whereas red denotes CNVs shared by two breeds.

Figure 1 .
Figure 1.The distribution of CNVs in various horse breeds.Breed-specific and shared CNV distribution throughout horse breeds.Breed-specific CNVs are shown in blue.Green denotes CNVs shared by three breeds, whereas red denotes CNVs shared by two breeds.

Figure 2 .
Figure 2. The number of CNVRs here identified and shared among the three horse breeds.

Figure 2 .
Figure 2. The number of CNVRs here identified and shared among the three horse breeds.

Genes 2023 , 15 Figure 3 .
Figure 3. Map of CNVRs in the horse genome.Circles and triangles represents DEL and DUP, respectively.

Figure 4 .
Figure 4. Correlation between chromosomal length and the number of CNVRs.

Figure 3 . 15 Figure 3 .
Figure 3. Map of CNVRs in the horse genome.Circles and triangles represents DEL and DUP, respectively.

Figure 4 .
Figure 4. Correlation between chromosomal length and the number of CNVRs.

Figure 4 .
Figure 4. Correlation between chromosomal length and the number of CNVRs.

Figure 5 .
Figure 5. Size range distribution of the CNVRs detected.

Figure 5 .
Figure 5. Size range distribution of the CNVRs detected.

Table 4 .
Analysis of KEGG pathways.

Table 1 .
The statistical results on CNV detection.
Breed Sample Number CNVRs Duplications Deletions Average Number of CNVRs Per Individual

Table 3 .
The descriptive statistics of the CNVRs on the autosomes of horses.

Table 4 .
Analysis of KEGG pathways.

Table 5 .
Analysis of GO pathway.
Figure 6.Functional classifications of the detected CNVRs.

Table 5 .
Analysis of GO pathway.