Genetic Diversity and Selective Signature in Dabieshan Cattle Revealed by Whole-Genome Resequencing

Simple Summary To protect the genetic resources of Chinese native cattle breeds, we investigated the genetic structure, genetic diversity and genetic signature from artificial or natural selection by sequencing 32 bovine genomes from the breeding farm of the Dabieshan population. We discovered that the ancestral contributions of Dabieshan originated from Chinese indicine and East Asian taurine on the autosomal genome, which had abundant genomic diversity. Some candidate genes associated with fertility, feed efficiency, fat deposition, immune response, heat resistance and the coat color were identified by a selective sweep. The SNPs data were based on genomics, which could establish a foundation for breed amelioration and support conservation for indigenous cattle breeds. Abstract Dabieshan cattle are a typical breed of southern Chinese cattle that have the characteristics of muscularity, excellent meat quality and tolerance to temperature and humidity. Based on 148 whole-genome data, our analysis disclosed the ancestry components of Dabieshan cattle with Chinese indicine (0.857) and East Asian taurine (0.139). The Dabieshan genome demonstrated a higher genomic diversity compared with the other eight populations, supported by the observed nucleotide diversity, linkage disequilibrium decay and runs of homozygosity. The candidate genes were detected by a selective sweep, which might relate to the fertility (GPX5, GPX6), feed efficiency (SLC2A5), immune response (IGLL1, BOLA-DQA2, BOLA-DQB), heat resistance (DnaJC1, DnaJC13, HSPA4), fat deposition (MLLT10) and the coat color (ASIP). We also identified the “East Asian taurine-like” segments in Dabieshan cattle, which might contribute to meat quality traits. The results revealed by the unique and valuable genomic data can build a foundation for the genetic improvement and conservation of genetic resources for indigenous cattle breeds.


Introduction
Cattle (Bos taurus) were amongst the first livestock species domesticated by humans [1]. Approximately 10,000 years ago, wild aurochs (Bos primigenius) began to be domesticated by humans [2], which were dispersed worldwide with the development of crops. Currently, more than 1000 breeds and varieties have been identified worldwide [3]. Based on habits and morphological characteristics, Bos taurus can be classified as two subspecies, B. t. taurus (humpless taurine) of European origin, and B. t. indicus (humped indicine) of Indian origin [4]. There are 55 domestic cattle breeds in China, which are valuable genetic resources to promote cattle breeding under changing production and environmental conditions in the future [5].
In our study, the experimental procedures were approved by the Experimental Animal Management Committee (EAMC) of Northwest A&F University (2011-31,101,684) and complied with the National Standard of Laboratory Animals Guidelines for Ethical Review of Animal Welfare (GB/T 35892-2018) and Guide for the Care and Use of Laboratory Animals: Eighth edition.

Genomic Diversity
In order to reduce the experimental error caused by the number of animals, we randomly selected 15 Dabieshan cattle and 15 Qinchuan cattle. The nucleotide diversity of each breed was carried out in 50 kb windows with 50 kb steps using VCFtools [19] (RRID:SCR 001235). Using PopLDdecay software [20] (RRID:SCR_022509) with default parameters, we estimated linkage disequilibrium (LD) decay with the physical distance between pairwise SNPs. The length and number of runs of homozygosity (ROH) per individual were calculated using VCFtools [19] with the following parameters: -homozyg-density 50homozyg-window-het 3-homozyg-window-missing 5. Based on four categories (0.5-1 Mb, 1-2 Mb, 2-4 Mb, >4 Mb), the ROH of each breed was classified.

Genome-Wide Selection
The selective sweep in genomics of Dabieshan cattle was calculated using the following strategy. First, the positive selection regions within Dabieshan cattle were computed by the nucleotide diversity (θπ) and the composite likelihood ratio (CLR) [21]. The θπ analyzed the diversity of the complete polymorphism data using VCFtools with 50 kb window size and 20 kb step size [19]. The CLR checks were estimated using SweepFinde2 [22] for sites within a non-overlapping 50 kb window. The top 1% windows of each method were considered as candidate regions. The Tajima's D statistic was performed by using VCFtools for the candidate gene.
Secondly, we performed comparisons between Dabieshan cattle versus Hanwoo using the fixation index (F ST ) [23], genetic diversity ratio (θπ-Ratio) [23] and cross-population extended haplotype homozygosity (XP-EHH) [24] for revealing genetic variants related to the adaptation of Dabieshan cattle. The F ST value, a method of statistical selective sweep regions based on population differentiation, is generally used to measure the reduction in heterozygosity of the target population relative to the reference population. θπ-ratio is the ratio of θπ values between the different populations of one species, and represents the degree of differentiation. We used VCFtools to perform F ST and θπ-ratio analysis [19] with 50 kb sliding and 20 kb step. For the XP-EHH, selscan v1.3.0 [25] was used to calculate the positive selection signals of each population in each 50 kb window. The top 1% (F ST > 0.58, θπ-ratio > 0.24, XP-EHH > 2.45) was chosen as the significance threshold, and genes in those regions were defined as potential candidate genes. We performed comparisons between Dabieshan cattle versus Qinchuan cattle using the F ST , which is consistent with the previous approach. The top 1% windows (F ST > 0.38) were deemed to be in the candidate region.
Finally, we applied RFMix v2.03-r0 [26] to perform East Asian taurine ancestry inference on autosomes of the Dabieshan cattle genome. We selected the genome of Hanwoo as the reference group in order to calculate "East Asian taurine-like" segments.

Functional Prediction Analysis
Gene ontology (GO) biological processes (BP) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were enriched using KOBAS 3.0 (http://bioinfo.org/kobas/, accessed on 5 March 2022) [27], which were significant when the corrected p-value was less than 0.05 for the understanding of the biological functions and signal transduction pathways of candidate genes in Dabieshan cattle.

Whole-Genome Sequencing, Assembly, and Genetic Variation
Thirty samples generated an~9.10 × coverage each (Table S1), and then two additional WGS data of Dabieshan cattle were generated from Chen et al. [28]. A total of 42,407,831 biallelic SNPs were identified in 32 Dabieshan cattle (Table S1). Figure 1a illustrates the functional classification of the polymorphic sites of 32 Dabieshan cattle. Most SNPs existed in intergenic (59.00%) and intron (38.05%) regions, and the remaining SNPs were present in the downstream (0.63%) and upstream (0.58%) regions of the open reading frame, untranslated regions (0.98%) and exonic regions (0.76%) that contained 120,972 non-synonymous SNPs and 192,147 synonymous SNPs.
In ordered to estimate the diversity of Dabieshan cattle, a tripartite combined genotyping of 32 Dabieshan cattle, the public genomes from 5 "core" cattle populations and Qinchuan cattle (Table S2) was performed using GATK. In total,~44 billion reads were generated, and the average alignment rate and sequencing depth reached 99.38% and~11.70×, respectively. The transition-to-transversion (Ts/Tv) ratios were 2.344-2.408 [29,30], indicating the good quality of our SNP data (Table S3). Furthermore, we performed a statistical analysis of SNPs and breed-specific SNPs based on the reference genome (ARS-UCD1.2). While comparing to other breeds, the number of SNPs and specific SNPs from taurine was significantly lower, which is consistent with previous studies [28]. Interestingly, the largest number (n = 42,407,831) and specific (n = 5,923,609) SNPs were found in the Dabieshan breed ( Figure S1), which might be partly explained by the different sample sizes per breed and partly by breed differences.

Population Structure
In order to explore the population structure of Dabieshan cattle, we assessed the neighbor-joining (NJ) tree, principal component analysis (PCA) and ADMIXTURE using the autosomal genome data. In PCA (Figure 1b), the first component separated Bos taurus from Bos indicus, explaining approximately 6.03% of the total genetic variation information, which indicated considerable genetic distance between taurine and zebu. In PC2, Chinese indicine, Indian indicine, East Asian taurine and European-Eurasian taurine were effectively separated, explaining approximately 2.37% of the total genetic variation information. The third component explained 1.80% of the total variation and could not separate any breed ( Figure S2). Moreover, the NJ tree ( Figure 1c) demonstrated that five "core" cattle populations formed their own separate clusters. Dabieshan cattle had the closest genetic distance from Chinese indicine, which is consistent with the PCA results. These results were also repeated by ADMIXTURE analysis [31] (Figure 1d). The results displayed that, at K = 2, 149, animals could be basically distinguished from indicine and taurine. At K = 4, Dabieshan cattle shared genome ancestry with Chinese indicine (0.857), East Asian taurine (0.139) and European taurine (0.004).

Patterns of Genomic Variation
The patterns of genomic variation in the Dabieshan population were based on the analysis of nucleotide diversity, linkage disequilibrium (LD) decay and runs of

Population Structure
In order to explore the population structure of Dabieshan cattle, we assessed the neighbor-joining (NJ) tree, principal component analysis (PCA) and ADMIXTURE using the autosomal genome data. In PCA (Figure 1b), the first component separated Bos taurus from Bos indicus, explaining approximately 6.03% of the total genetic variation information, which indicated considerable genetic distance between taurine and zebu. In PC2, Chinese indicine, Indian indicine, East Asian taurine and European-Eurasian taurine were effectively separated, explaining approximately 2.37% of the total genetic variation information. The third component explained 1.80% of the total variation and could not separate any breed ( Figure S2). Moreover, the NJ tree ( Figure 1c) demonstrated that five "core" cattle populations formed their own separate clusters. Dabieshan cattle had the closest genetic distance from Chinese indicine, which is consistent with the PCA results. These results were also repeated by ADMIXTURE analysis [31] (Figure 1d). The results displayed that, at K = 2, 149, animals could be basically distinguished from indicine and taurine. At K = 4, Dabieshan cattle shared genome ancestry with Chinese indicine (0.857), East Asian taurine (0.139) and European taurine (0.004).

Patterns of Genomic Variation
The patterns of genomic variation in the Dabieshan population were based on the analysis of nucleotide diversity, linkage disequilibrium (LD) decay and runs of homozygosity (ROH) with a window scale of 50 kb. In Figure 2a, we observed that the average LD (r 2 ) was the highest in Jersey cattle, followed by Hereford cattle, Simmental cattle and Angus cattle, when the distance between adjacent SNPs was short (<10 kb), and, on the other hand, that the lowest average LD was found for Dabieshan cattle, then Qinchuan cattle and then followed by Chinese indicine and Indian indicine. The nucleotide diversity of Dabieshan (mean θπ = 0.0034) was lower than that of Chinese indicine cattle (0.0037) but approximately three times higher than that of European breeds (0.0007-0.0010). The tendency of nucleotide diversity (Figure 2b) was nearly consistent with those from LD decay, in which, the highest genome-wide genetic diversity was found in indicine cattle (Indian and Chinese indicine) and Dabieshan cattle.
Biology 2022, 11, x 6 of 15 homozygosity (ROH) with a window scale of 50 kb. In Figure 2a, we observed that the average LD (r 2 ) was the highest in Jersey cattle, followed by Hereford cattle, Simmental cattle and Angus cattle, when the distance between adjacent SNPs was short (<10 kb), and, on the other hand, that the lowest average LD was found for Dabieshan cattle, then Qinchuan cattle and then followed by Chinese indicine and Indian indicine. The nucleotide diversity of Dabieshan (mean θπ = 0.0034) was lower than that of Chinese indicine cattle (0.0037) but approximately three times higher than that of European breeds (0.0007-0.0010). The tendency of nucleotide diversity (Figure 2b) was nearly consistent with those from LD decay, in which, the highest genome-wide genetic diversity was found in indicine cattle (Indian and Chinese indicine) and Dabieshan cattle.  breeds. Figure 2c shows that the majority of ROHs were found between 0.5-1 Mb, whereas, sensibly, European commercial breeds had more medium-sized (2-4 Mb) and long ROHs (>4 Mb). In addition, Dabieshan cattle presented a lower length of ROH compared to European commercial breeds, which had a greater length (Figure 2d). Consanguineous mating leads to the generation of long ROH, whereas shorter ROH mirror distant ancestral influences [32]. As expected, neglecting the sample size, the number and length of ROH were generally higher in taurine than in hybrid and indicine, suggesting that commercial breeds might have a risk of a decline in inbreeding.

Candidate Genes under Selective Sweep
The nucleotide diversity analysis (θπ) and the composite likelihood ratio (CLR) were employed to distinguish the positive selection region in the Dabieshan breed alone. Outlier signals (top 1%) were considered as candidate selective regions (Figure 3a). Based on the θπ test statistic, we obtained 1103 putative genes (Table S4), whereas 529 positively selected genes were detected by the CLR test (Table S5). These genes included GPX5, GPX6 (reproductive performance) [33], SLC2A5 (feed efficiency) [34][35][36], IGLL1 [37] and BOLA-DQA2, BOLA-DQB (immunity) [8,[38][39][40]. It is worth noting that the MLLT10 gene, which is associated with fat deposition [41], was detected in both the two methods, indicating that the mentioned genes were strongly selected in Dabieshan cattle. The positive selection signals around the region of MLLT10 (BTA13: 22.74-22.95 Mb) were further recapitulated by significantly lower values of nucleotide diversity and Tajima's D in Dabieshan cattle (Figure 3b). To better understand the biological implications of the selection signals, we performed functional enrichment analysis for genes with selection signatures ( Figure S3). A gene set enrichment analysis of KEGG pathways revealed that the most significant pathway was the "thyroid hormone signaling pathway" (corrected p value = 0.004, Table S6), which is related to heat resistance. Over-representation analysis of GO BP terms showed that Dabieshan cattle had increased GO categories involved in "embryonic skeletal system morphogenesis" (corrected p value = 2.467 × 10 13 ) and "ATP binding" (corrected p value = 1.202 × 10 6 ) ( Table S7).
We divided the genome into non-overlapping segments of 50 kb [42] to compare the genomes of Dabieshan cattle to distinguish between East Asian taurine (Hanwoo) signatures of positive selection following environmental and human selection pressures. Outlier regions (the top 1% F ST , θπ ratio or XP-EHH statistics) were deemed to be candidate regions for further analysis. Based on F ST , we obtained 884 candidate genes (Table S8). In contrast, 757 and 600 candidate genes were identified using the θπ ratio and XP-EHH (Tables S9 and S10) separately. Some of the candidate genes were outstanding genes related to heat resistance (DnaJC1, DnaJC13, HSPA4) [43][44][45][46], so they had a strong positive selection with respect to the adaptive characteristics of Dabieshan cattle (Figure 3c). By calculating F ST with a smaller window (5 kb) and haplotype patterns, significant differentiation was observed between Dabieshan cattle and Hanwoo (Figure 3d). Synchronously, a missense mutation (p.Pro522Leu) was observed in the heat-resistance-related gene DnaJC1 (Figure 3e). Allele G displayed a rare distribution (frequency 0.172) in Dabieshan cattle, whereas it showed an opposite pattern (frequency 1.000) in Hanwoo. Therein, the HSPA4 gene was overlapped by three analyses. We also found one missense mutation that encodes a p.Leu792Pro substitution, and that the homozygous mutation (CC) and heterozygous mutation (CT) were highly conserved in the Dabieshan population (0.97) ( Figure S4). Moreover, we noticed a region (25.58-25.63 Mb) scanned by the θπ ratio on chromosome 23 (including BOLA-DQA2 and BOLA-DQB), illustrating a strong positive selection signal (Figure 3c). In the functional prediction of all candidate genes, significant KEGG pathways and GO BP terms (corrected p < 0.05) were mainly enriched in immunity ("immune response, GO:0006955"), heat resistance ("adrenergic signaling in cardiomyocytes, bta04261", "thyroid hormone signaling pathway, bta04919") and coat color ("melanogenesis, bta04916") (Tables S11 and S12 and Figure S5). The coat color of 32 Dabieshan cattle was black or gray-black. We tried to investigate the black or gray-black color of Dabieshan cattle by comparing them with Qinchuan cattle (reddish-brown). A significant differentiation of the ASIP gene was identified (F ST = 0.4695), which was further verified by haplotype patterns ( Figure S6). Previous research has suggested that the variant at the ASIP locus causes coat color darkening in Nellore cattle [47], which could be linked to the coat color variation in Dabieshan cattle. ("melanogenesis, bta04916") (Tables S11 and S12 and Figure S5). The coat color of 32 Dabieshan cattle was black or gray-black. We tried to investigate the black or gray-black color of Dabieshan cattle by comparing them with Qinchuan cattle (reddish-brown). A significant differentiation of the ASIP gene was identified (FST = 0.4695), which was further verified by haplotype patterns ( Figure S6). Previous research has suggested that the variant at the ASIP locus causes coat color darkening in Nellore cattle [47], which could be linked to the coat color variation in Dabieshan cattle.  Finally, we calculated 6,967,850 "East Asian taurine-like" segments in Dabieshan cattle, and a total of 1207 genes were identified in high-frequency regions (>0.9) (Figure 4a, Table S13). Notably, from the DAVID gene ontology, there were some significant GO BP pathways related to muscle and skeleton development, such as "actin cytoskeleton, GO:0015629", "actin filament network formation, GO:0051639", "calcium ion binding, GO:0005509" and "BMP receptor binding, GO:0070700" (Figure 4b, Table S14).
Biology 2022, 11, x 9 of 15 haplotype patterns heatmap and structure of DNAJC1 with exons indicated by vertical bars. A missense mutation was highlighted in red.

Discussion
Since the 1980s, a large number of exotic commercial cattle breeds have been introduced to blindly improve Chinese local cattle, leading to the disintegration of the excellent gene pool due to crossbreeding, and the genetic resources of Chinese local cattle are gradually being depleted. In order to protect the precious genetic resources of Chinese native cattle, it is crucial to design scientific and practical breeding programs that can contribute to the conservation of local Chinese cattle. Genomic data could reflect a wide variety of historical developing events, including climate adaptation, species introductions and artificial selection [48]. Thus, the exploration of population genetic structure and genetic

Discussion
Since the 1980s, a large number of exotic commercial cattle breeds have been introduced to blindly improve Chinese local cattle, leading to the disintegration of the excellent gene pool due to crossbreeding, and the genetic resources of Chinese local cattle are gradually being depleted. In order to protect the precious genetic resources of Chinese native cattle, it is crucial to design scientific and practical breeding programs that can contribute to the conservation of local Chinese cattle. Genomic data could reflect a wide variety of historical developing events, including climate adaptation, species introductions and artificial selection [48]. Thus, the exploration of population genetic structure and genetic diversity can be regarded as essential for genetic evaluation, reflecting recent cross-breeding and the utilization or conservation of genetic resources. The evaluation of population genetic structure and genetic diversity using mitochondrial DNA and Y chromosome markers is less precise than the assessment from the WGS data; therefore, the data obtained by WGS were adopted in this experiment.
A previous WGS analysis of 49 worldwide cattle breeds suggested that the domesticated cattle were divided into five "core" populations: Chinese indicine, Indian indicine, East Asian taurine, European taurine and Eurasian indicine [28]. In this study, we employed the five "core" cattle populations and an outstanding Chinese local breed as a reference group to explore the population genetic structure and genetic diversity of Dabieshan cattle. Dabieshan cattle are hybrid cattle, the ancestral contributions of which originated from Chinese indicine (85.7%) and East Asian taurine (13.9%), as shown in the ADMIXTURE analysis (Figure 1d). The nucleotide diversity, LD decay and ROH results are consistent with prior studies [49], directly indicating that the genetic diversity was higher and that inbreeding depression was lower compared to European commercial varieties. Chinese indicine had a high nucleotide diversity on the autosomal genome, which may be a prior reception of introgression from Banteng (Bos javanicus) [28]. The relatively abundant genomic diversity was found in Dabieshan cattle, which may result from hybridization with taurine and indicine. At the same time, populations of Dabieshan and Chinese indicine inhabit small farms and undergo constant selection and elimination with less inbreeding. The monotonous feeding environment and high degree of inbreeding may be responsible for the lower nucleotide diversity of European commercial breeds than indicine and hybrid cattle [50]. However, deleterious variants might reach a high frequency in European commercial breeds due to the frequent utilization of individual carrier animals in artificial insemination [51] or hitchhiking with favorable alleles under artificial selection [52,53], and even due to missing superior genes from stress-resistance traits such as disease resistance and heat stress tolerance.
In comparison to SNP arrays, we were able to identify more selective signatures using WGS data [54,55]. Using complementary approaches, we discerned several candidate regions for superior traits that were potential targets of artificial and natural selection in Dabieshan cattle.
Considering that Dabieshan cattle express a better performance in adapting to a hightemperature and high-humidity environment [5], further comparisons were made between Dabieshan and commercial breeds (Hanwoo). The DnaJC1 gene (F ST = 0.86) translates an endoplasmic reticulum heat shock protein that binds the molecular chaperone HSPA5 (alias BiP) [43]. The DnaCJ13 gene-encoded protein, playing a role in clathrin-mediated endocytosis, is associated with the heat-shock protein Hsc70 [44]. The heat-shock protein family A (Hsp70) member 4 (HSPA4) is notable for promoting cell protection against heat damage and preventing protein denaturation [45,46]. Our KEGG analysis revealed two significant heat-stress-related pathways ("adrenergic signaling in cardiomyocytes" and "thyroid hormone signaling pathway") ( Figure S5). The epinephrine was elevated with prolonged environmental heat stress [56]. Meanwhile, thyroid hormones adjust the metabolic rates to facilitate the body thermal balance, which is quite important for heat adaption [57]. These results might have rapidly evolved in Dabieshan cattle, which might give a reasonable explanation for their utterly different degree of thermotolerance compared to the commercial breed. Dabieshan cattle are generally more resistant to disease than their commercial counterparts [5]. Many of our candidate regions harbored genes implicated in the immune system, and major histocompatibility complex (MHC) genes were annotated (BoLA, BOLA, BOLA-DQB, BOLA-DQA2 and BOLA-DQA5). The MHC region contains a diverse array of genes that are crucial for the initiation of adaptive immune responses that map to chromosome 23 in cattle [38]. Notably, extensive studies have reported the importance of the bovine lymphocyte antigen complex in host immunity [39] and its correlation with parasitic diseases [8,40]. The GBP4 and GBP6 gene were detected in θπ and the θπ-ratio. GBPs mediate resistance against invading pathogens [58]. In the functional enrichment of all candidate genes, the significant "immune response, GO:0006955" pathway (corrected p = 0.01) was enriched (Table S12 and Figure S5).
Bovine fecundity is a crucial factor for a successful reproductive performance; as for Dabieshan cattle, they possess early sexual maturation (10-13 months) [5]. In the glutathione peroxidases (GPxs) gene family, the GPX5 protein plays a vital role in the maintenance of spermatozoa DNA integrity [59]. The GPX6 is regarded as an essential participator in the in vitro induced capacitation and acrosome reaction in porcine sperm [33]. Feed efficiency is an important economic factor affecting beef production costs [60]. It has been reported that Dabieshan cattle have the characteristic of roughage resistance and have a high potential to produce highly marbled beef [5]. Similarly, Dabieshan cattle as tillage livestock have the typical characteristics of muscularity and roborant skeletons, with intense artificial selection. A large amount of evidence shows that the SLC2A5 gene is involved in the metabolism of carbohydrates [34], and the increased expression of SLC2A5 has been associated with fat deposition [35]. The increased SLC2A5 expression was associated with an enhanced relative growth rate and Kleiber index in Nelore steers [36]. The copy number variation of the MLLT10 gene sensibly influenced the growth traits of two Chinese native cattle breeds [41]. Similarly, the MLLT10 gene had a feed conversion ratio (FCR) regarding male Duroc pigs [61]. Furthermore, the functions of these KEGG and GO pathways from the "East Asian taurine-like" segments may contribute to fat deposition and muscle and skeletal development (Figure 4b). Based on GeneCards annotation (https://www.genecards.org/, accessed on 10 May 2022), the ROCK2 gene regulates the formation of actin stress fibers, which is related to intramuscular fat under selection in Ankole cattle [62]. Through the activation of the mTOR signaling pathway, an increase in the level of bone morphogenetic protein 7 (BMP7) transcription or activity of its receptor in muscle cells can induce hypertrophy [63]. On the contrary, the BMP4 growth factor increases the number of activated satellite cells and regulates muscle regeneration [64,65]. Moreover, the PPP3R1 gene encodes a phosphatase protein, and the expression levels of the PPP3R1 gene may result in differences in muscle deposition in Nelore cattle [66]. The PPP3CA gene has been shown to be involved in the differentiation of perimuscular preadipocytes in cattle [67], which are responsible for the storage of fat [68]. A famous beef meat grade, A4 (the Japanese Beef Meat Grade scale, 2008), can be produced by purebred Dabieshan cattle with nutritional regulation. Grade A4 is recommended as the most highly satisfactory and most economical top grade of fattened beef for Chinese consumers [69], which have tremendous market potential. Because of these genes and pathways, we hypothesize that Dabieshan cattle have the potential to become a source of high-quality beef production.
Thirty-two Dabieshan cattle from the core breeding farm have purified and uniform ancestral contributions with a small impact of globalization, which is suitable for breed conservation and the in-depth exploration of genetic resources. We highlighted some adaptations at the genome level, which may relate to climatic challenges (e.g., heat and disease) and artificial selection (e.g., fertility, fat deposition, fat deposition, coat color). Some of our results might provide deep insight toward enhancing livestock production and increasing adaptation to local environmental challenges through crossbreeding (e.g., marker-assisted selection to increase haplotype frequency). In practice, these results may develop new genomic evidence for the design and implementation of improved poultry productivity. The Dabieshan cattle are predominantly fed in small groups (approximately 10-20) by farmer. Small group feeding affects a high degree of inbreeding, which can lead to the inhibition of various crucial traits. At the same time, the reproduction rate and productivity are unstable due to extensive feeding. The breeding efficiency is low under the traditional breeding mode; the breeding stock gradually declines. At present, there are many native cattle breeds in the world that have excellent traits, and not only Dabieshan cattle, but they are in a state of scattered breeding, and the sufficient protection and utilization of their genetic resources need to be investigated.

Conclusions
Our study presented a comprehensive overview of the Dabieshan genome by applying whole genome sequencing. Dabieshan cattle originated from Chinese indicine and East Asian taurine and demonstrated abundant genomic diversity. Additionally, the highlighted regions and genes that provide evidence of past and ongoing selection in Dabieshan cattle are enriched for candidate genes that may be associated with immune response, heat stress, reproduction, feed efficiency and meat quality traits. These results extend the application for further studies on the genomic characteristics, genetic resource evaluation and conservation strategies for Chinese native cattle breeds.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/biology11091327/s1, Figure S1: Venn diagram showing overlapping and unique SNPs among different populations in the reference genome (ARS-UCD1.2); Figure S2: Principal component analysis, PC1 against PC3; Figure S3: GO and KEGG pathways from the enrichment analysis of genes that were selected by θπ and CLR; Figure S4: Structure of the HSPA4 gene with exons indicated by vertical bars. A non-synonymous SNP (rs210913195) represents p.Leu792Pro and is marked in red. Haplotype frequencies are indicated by numbers next to each haplotype. In each haplotype, green and yellow bars represent alleles 1 and 2, respectively; Figure S5: GO and KEGG pathways from the enrichment analysis of genes that were selected by F ST , θπ-ratio and XP-EHH; Figure S6: Selective signals between Dabieshan cattle and Qinchuan cattle. Manhattan plot of selective sweeps by F ST methods and haplotype diversity of the ASIP gene; Table S1: Summary of the sequencing data; Table S2: List of additional cattle samples for analysis of genetic background in Dabieshan cattle; Table S3: Distribution of SNPs annotation in cattle breeds (the sample size larger than 5); Table S4: Summary of genes from θπ in Dabieshan cattle; Table S5: Summary of genes from CLR in Dabieshan cattle; Table S6: KEGG pathway analysis of Dabieshan candidate genes by θπ and CLR methods; Table S7: GO pathway analysis of Dabieshan candidate genes by θπ and CLR methods; Table S8: Summary of genes from F ST ; Table S9: Summary of genes from θπ-ratio; Table S10: Summary of genes from XP-EHH; Table S11: KEGG pathway analysis of Dabieshan candidate genes by F ST , θπ-ratio and XP-EHH; Table S12: GO pathway analysis of Dabieshan candidate genes by F ST , θπ-ratio and XP-EHH; Table S13: Summary of genes from "East Asian taurine-like" segments in Dabieshan cattle; Table S14: GO pathway analysis of Dabieshan candidate genes from "East Asian taurine-like" segments.
Author Contributions: N.C. and X.G. conceived and designed the experiments. X.G., S.Z. and W.X. performed the statistical analysis. S.Z., H.J. and Y.J. performed the sample DNA extraction and data upload. N.C. provided technical assistance. S.Z., C.L., N.C., H.J. and Y.J. contributed to the sample collections. C.L., N.C. and L.X. provided the laboratories for DNA extraction and statistical analysis. X.G. and W.X. drafted the manuscript. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement:
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.