Decoding the beef cattle genome through whole-genome sequencing reveals the genetic basis of carcass and meat quality traits

Abstract

Meat quality and carcass traits are the most economically important traits affecting the value and the quality characteristics of the animal at slaughter. In this study, we analyzed the genomes of 295 domestic cattle from 20 global breeds using next-generation sequencing to identify signatures of positive selection associated with meat production and quality traits. The results of population genetic analysis revealed genetic differentiation among worldwide cattle groups. We observed clustering of samples in agreement with their geographic origins and production characteristics. We further observed noticeable genetic variety between cattle groups from different geographical regions. Generally, the lowest genetic diversity was determined for commercial cattle breeds, while the highest genetic diversity was found in African local cattle breeds. Our search for putative selective genomic regions in beef cattle populations revealed several candidate genes such as BMP2, EGR1, MAGEL2, U6, TMEM201, ELK3 and 5S_rRNA that were previously explored to be associated with carcass traits and meat quality in beef cattle. The enriched pathways and candidate genes discovered in this study could supply a basis for future improvement through the use of whole-genome technologies and selective breeding.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-04205-y.

Introduction

The growth perspective of the global demand for animal-based protein products is expected to be up around 50% until 2050, resulting in an increased demand to enhance productivity, efficiency, and management^1,2. In many parts of the globe, livestock plays important roles in farming systems. They are an important component of the agricultural sector, which supplies a wide variety of high-quality products^3,4. Beef is one of the most common sources of animal protein containing the full range of necessary amino acids needed in a human diet⁵. Beef quality and quantity can be influenced by different factors such as; breed, feed management, carcass aging, and animal health⁶. For many beef producers, improving the quality of the meat is a well-established breeding objective. For example, marbling, which refers to the flecks of fat that appear within the muscle tissue of beef, is one of the main criteria for evaluating the quality of beef. This factor affects the quality and grade of beef and has an important impact on beef producer decisions to sell their animals based on grade and yield^7,8.

During the past decades, scientists have made more efforts to provide a better insight into the genetic basis related to economically important traits. On a global scale, genetic selection for meat-related traits is crucial for enhancing the efficiency of beef production and is likely key to the sustainability of all meat-producing livestock species^9,10. Beef breeders have been focused on improving major carcass traits such as carcass weight, ribeye area, yield quality grade, fat thickness and yield grade.

Previous genome-scan studies that employed techniques such as marker-QTL linkage analysis^11,12, candidate gene approaches¹³, genome-wide association studies (GWAS)^14–16 and whole-genome sequencing (WGS) analysis^17,18 have reported genomic regions and candidate genes related to carcass traits in different beef cattle breeds. However, only a few genomic studies focusing on carcass traits have included multiple beef cattle breeds. In this study, whole genome sequence data from worldwide cattle breeds were utilized to characterize the population structure, genetic diversity, and signatures of selection analysis. We discovered novel genomic regions that likely play a role in the carcass and meat quality traits and may be under selection in the studied populations. This study could expand the understanding of the genetic architecture behind key traits, offering new targets for selection in beef cattle breeding objectives.

Results

Genetic diversity and population structure

We performed different classical analyses to investigate the genetic population structure and relatedness of the worldwide cattle groups. A maximum likelihood (ML) phylogenetic tree was constructed to infer the genomic relationships of all studied cattle. Based on this tree (Fig. 1A), all commercial cattle breeds were first clustered into two phylogenetic clades and then separated from other local cattle individuals. A higher degree of genetic relatedness between some cattle breeds, such as Angus (ANG), Shorthorn (STH), Limousin (LMS), and Charolais (CLR), may reflect a greater similarity in the genetic makeup of these breeds. These results were further supported by the PCA and ADMIXTURE analysis (Fig. 1B,C). A clear separation between commercial cattle breeds and other indigenous samples was observed in the first component (PC1) which is consistent with the phylogenetic tree (Fig. 1B). The PC1 and PC2, accounted cumulatively for ~ 62.5% of the total variation (52.29% and 10.25%, respectively) as shown in Fig. 1B. PC1 represented a genetic cline from the Indian local cattle group at the eastern end to European commercial beef cattle at the western end. Furthermore, among the local cattle breeds, Hanwoo (HNW) which is a breed of cattle indigenous to Korea and has been intensively bred for meat-related traits over the last decades¹⁹, presented a closer genetic distance to the western cattle groups (Fig. 1B). In contrast, PC2 corresponded to the relatively high genetic variation between Hereford (HRF) samples and other cattle groups. Local cattle groups could further be grouped into two distinct groups, the African group and other Asian populations. The cluster analysis implemented in ADMIXTURE (from K2 to K5) distinguished the ancestry component for all cattle samples that were examined (Fig. 1C). The K = 2 clearly splits all indigenous cattle samples of African and Asian groups from other cattle populations. In line with the PCA analysis, the K = 3, with the lowest cross validation (CV) error (CV = ~ 0.52), distinguishes samples from HRF group with other local and commercial cattle breeds. When K = 4, we observed a clear separation between ANG and STH populations and samples from other commercial cattle breeds such as Piedmontese (PIT), LMS, Limonero (LMN), Kholmogory (KMG) and CRL (Fig. 1C). Ancestral proportions at K = 5, divided the Indian zebu cattle group (GIR) from the remaining indigenous cattle samples from Africa.

Fig. 1.

Population genetic analysis among worldwide cattle breeds. (A) Phylogenetic tree constructed by FastTree analysis; (B) Principal component (PC) analysis, PC 1 against PC 2; (C) Model-based ancestry for each individual assuming admixture. Cattle breed abbreviations: PIT Piedmontese, CRL Charolais, ANG Angus, STH Shorthorn, HRF Hereford, HNW Hanwoo, LMS Limousin, ARS Arsi, BGT Begait, BTN Butana, FGR Fogera, GIR Gir, GFA Goffa, HRO Horro, IRN Iran, KNN Kenana, KMG Kholmogory, LQG Leiqiong, LMN Limonero, OGD Ogaden.

We further estimated the genome-wide distribution of nucleotide diversity (θπ) within each studied population. Our results revealed that the majority of native cattle breeds (especially African local breeds) have higher genomic diversity than other commercial groups (Fig. 2A). Additionally, to investigate the patterns of genetic divergence as a function of θπ, we grouped all cattle samples into three different groups according to their carcass traits and meat quality values (see Fig. 2B legend). The marbling group included cattle breeds that typically have a high amount of intramuscular fat (IMF) in their muscle tissue²⁰. The beef cattle group included breeds that could still have suitable meat quality but might not have the richness and tenderness associated with higher marbling, and the local cattle group consisted of global native breeds that have developed and adapted over time within specific geographical regions and are well-suited to their environmental conditions²¹. By comparing the genetic sequences of different groups, we found a positive relationship of nucleotide diversity measures between cow groups with the high marbling and beef production capacities, which also could be due to their shared genetic ancestry. In contrast, we observed a relatively lower correlation between the local cattle group and both beef and marbling cattle populations (Fig. 2B).

Fig. 2.

(A) Nucleotide diversity (θπ) of cattle populations; (B) Correlation of θπ between local cattle group (green), beef cattle group (blue) and marbling group (red) (50-kb non-overlapping window). For genomic correlation analysis, we divided the worldwide cattle populations into three groups. Marbling group included ANG, STH, HNW, and LMN breeds. Beef cattle group, including CRL, HRF, and PIT breeds.

Detection of highly differentiated genomic regions

Two different statistical approaches (FST and log2 θπ ratio) were applied to investigate genomic regions that show evidence of positive selection between the local cattle group and the beef cattle population. According to our comparative genomic analysis, a total of 208 and 173 protein-coding genes were found in outlier windows with significantly high FST (empirical P ≤ 0.01) and log2 θπ ratio scores (empirical P ≤ 0.01), respectively (Supplementary Tables S2 and S3, Fig. 3). All genomic regions discovered by the aforementioned methods were then annotated to the corresponding loci and biological pathways to explore the potential genetic mechanisms involved with various characteristics in beef cattle. The overlapping candidate gene results from both methods identified several loci (such as BMP2, EGR1, MAGEL2, U6, TMEM201, ELK3 and 5S_rRNA) previously associated with carcass traits in beef cattle (Table 1). Moreover, we used the g: Profiler web toolset for the gene ontology analyses. Our results revealed some categories involved with both growth and development of the skeletal system in cattle such as “Growth” (GO:0040007), “Animal organ development” (GO:0048513), “Limb development” (GO:0060173) and “developmental process” (GO:0048869) (Supplementary Table S4).

Fig. 3.

Manhattan plot of genome-wide FST values for differentiation between populations, candidate genes related to carcass traits are highlighted.

Table 1.

Overlapped candidate genes identified by two statistical methods (FST and log2 Θπ ratio) affecting carcass traits in beef cattle group.

Gene name	Chromosome	Genomic region (windows)	Summary of gene function
5 S-rRNA	BTA*-7	46,325,000–46,375,000	Related with carcass traits52
BMP2	BTA-13	49,150,000–49,200,000	Body size traits44–47
DVL1	BTA-16	51,175,000–51,225,000	Carcass traits in beef cattle37–39
EGR1	BTA-7	49,775,000–49,925,000	Carcass traits, adipogenesis72,73
ELK3	BTA-5	60,450,000–60,500,000	Carcass traits41–43
FANCA	BTA-18	14,625,000–14,675,000	Growth and carcass quality traits74,75
HOXD10	BTA-2	20,750,000–20,800,000	Meat quality traits76
LYN	BTA-14	23,125,000–23,175,000	Associated with feed intake and growth in beef cattle75,77
MAGEL2	BTA-21	1,200,000–1,250,000	Candidate gene affecting carcass traits78
U6	BTA-15	43,075,000–43,125,000	Carcass traits52

*Bos taurus autosome.

Discussion

Genome diversity and relationships

Understanding the genetic basis of diversity and population structure within and among global cattle populations is essential for developing effective conservation strategies and advancing sustainable breeding practices^22,23. In this study, we examine genetic variation patterns in several cattle populations by using complete genome sequence data from more than 290 cattle individuals. Consistent with the majority of earlier studies, we observed that the local cattle groups from different geographical regions harbor higher levels of genetic diversity compared with commercial cattle breeds, which might be because of traditional breeding practices applied by local farmers and also random mating^4,24,25. Lower genetic diversity in commercial cattle breeds could be due to artificial selection for desired traits such as carcass and meat production traits^26,27. Among the studied populations, we observed lower levels of nucleotide diversity for HRF and ANG, respectively, which is in agreement with previous studies and could be attributed to their intense artificial selection during the last decades^28,29. The findings from different population genetic analysis, such as PCA, admixture (K = 2 to K = 5), and phylogeny, have collectively shown shared genetic ancestry between European beef cattle breeds (except the HRF breed). Moreover, our results provide compelling evidence of shared genetic ancestry among local African cattle breeds, consistent with previous findings^24,25.

Genomic regions related to meat production traits

The profitability of beef cattle farming is impacted by different key components that directly influence carcass value, such as carcass weight, carcass conformation, and carcass fat. Given the economic significance of carcass merit to producers, it remains a primary objective in beef breeding^30,31. Previous studies have shown that there is substantial genetic variation in growth and carcass traits in different cattle populations^32–34. Therefore, a thorough comprehension of the genetic variants that are involved in carcass merit is useful to maximize the efficiency of breeding for genetic improvement of carcass traits. Over the past two decades, numerous GWAS studies have provided valuable insights into the genetic architecture underlying beef carcass traits, helping to enhance our understanding of the molecular mechanisms involved in these traits^16,35,36. Several significant genes and genomic regions have been discovered that contribute to key beef carcass traits. Our findings from the identification of selection signatures between beef cattle and local cattle populations discovered a number of potential candidate genes in the regions of high confidence selection (top 1% F_ST and log2 θπ ratio values) that could be associated with carcass and meat quality traits (Table 1, Supplementary Tables S1 and S2). In our broad spectrum analysis, several candidate genes such as; 5S_rRNA, BMP2, DVL1, EGR1, ELK3, FANCA, HOXD10, LYN, MAGEL2, U6 and TMEM201 located on different chromosomes were found to be most likely associated with carcass traits in beef cattle (Table 1; Fig. 3).

Dishevelled homolog 1 (DVL1) gene was found as one of the candidates on BTA16. The specific expression of this gene in skeletal muscle tissue and its important function in controlling muscle development have been reported in different species^37,38. A GWAS study that analyzed data from more than three hundred cattle individuals identified that the SNP-ss96319521 located within this gene could be considered a desirable candidate marker for carcass traits in beef cattle. It has been reported that individuals carrying the CC genotype have significantly higher levels of both rib-eye muscle area and carcass weight at slaughter³⁹. Another possible candidate gene, ELK3, was observed in one of the selection regions on BTA5. This gene encodes a transcription factor that plays a significant role in regulating various cellular processes by binding to specific DNA sequences⁴⁰. This ability to influence gene expression suggests that it may indirectly affect metabolic pathways, including those involved in carcass and beef traits. The association of this gene with important traits in beef cattle, such as rib-eye area⁴¹, meat quality⁴² and daily weight gain⁴³ has been confirmed in previous studies. In addition, we identified the bone morphogenetic protein 2 (BMP-2) gene on BTA13, which is a multifunctional growth factor belonging to the transforming growth factor-beta (TGF-β) superfamily. Previous genomic analysis on different species, such as human⁴⁴, sheep^45,46, pig⁴⁷ and horse⁴⁸ have found associations of the BMP2 gene variants with body size traits. This gene has been identified as a key regulator of bone growth and development traits in pigs⁴⁷, which are considered economically important traits for meat production. Furthermore, BMP2 signaling has been linked to intramuscular fat (IMF) content, affecting meat quality and yield in cattle. It has been proposed that BMP2 stimulates both hyperplasia and hypertrophy in bovine preadipocytes via activating the BMP/SMAD signaling pathway⁴⁹. Additionally, the gene expression of this protein is up-regulated in response to castration which led to increased adipogenic gene expression and also higher IMF content⁵⁰. The 5 S-rRNA is another candidate gene that is associated with carcass traits in cattle. Wang et al.⁵¹, based on imputed whole genome sequence data from multiple beef cattle breeds, reported that 5 S-rRNA could be considered as one promising candidate gene for carcass merit traits. A previous study integrating GWAS with copy number variant (CNV) and SNP data identified several candidate genes related to carcass traits in cattle. Notably, multiple copies of the 5 S-rRNA gene were reported in close proximity to CNVs and SNPs associated with carcass traits, such as carcass fat and conformation⁵². This suggests a potential role for the 5 S-rRNA gene in influencing carcass-related traits. There are also a number of studies that have utilized candidate gene approaches and GWAS to identify genomic regions that harbor genes influencing carcass and beef related traits^16,35,36. For example, several different genes such as EDGPR1, Titin, Akirin 2, and RPL27 were considered as functional candidates for the marbling trait⁵³. Furthermore, it has been demonstrated that the expression of carcass traits such carcass weight, fat thickness, rib-eye area, and meat yield is significantly influenced by genetic diversity^54,55. Recent studies highlight the importance of preserving genetic diversity while selecting for these traits, as it is crucial for the long-term sustainability of beef production systems⁵⁶.

Nonetheless, it is clear that the genetic makeup of animals, particularly at the breed or population level, is influenced not only by genes associated with desirable traits (such as production traits) but also by genes that contribute to adaptation, disease resistance, and other important characteristics. For example, cattle in tropical regions have genes that promote better sweat gland function, skin pigmentation, and other mechanisms to enhance heat dissipation⁵⁷. Understanding the genetics behind these traits is also a key element in developing more sustainable and productive cattle breeds. In this study, some candidate genes with higher signature values (p value < 0.01) were also found to be related to climate adaptations and disease resistance traits (Tables S3-S4). For example, we identified DNAJC11 and DnaJC1 genes, which are related to heat resistance traits in local cattle^58,59. Some other genes, such as TLR3⁶⁰ and TRPS1⁶¹, are associated with disease-resistant traits.

In conclusion, this study provides strong evidence of significant genetic divergence between local cattle populations and commercial breeds, which are widely recognized for traits related to beef production. Nonetheless, we found no significant genetic differentiation between beef cattle groups, particularly between the LMS, CRL, STH, and ANG breeds. Furthermore, by applying two different statistical approaches, we identified several candidate genes involved in carcass and meat quality traits. Our findings contribute to a deeper understanding of the genomic mechanisms underlying beef cattle production. By identifying the targets of selection in populations bred for enhanced meat yield and related traits, this research paves the way for future genome-wide association studies and further investigations into genomic selection targets in beef cattle.

Materials and methods

Ethics statements

As no biopsies and blood samples were obtained from living animals, no ethics statement was required for this study. The raw sequence data used in this study were downloaded from the National Center for Biotechnology Information (NCBI) GenBank, as it was a study based on the use of whole genome sequencing data (Supplementary Table S1).

Datasets, quality checking and SNP calling

In this study, publicly available genomic data were obtained from NCBI SRA Database. We collected worldwide complete genome sequences from 20 different cattle breeds including; PIT (n = 10), CRL (n = 26), ANG (n = 25), STH (n = 25), HRF (n = 20), HNW (n = 23), LMS (n = 25), ARS (n = 10), Begait (BGT; n = 9), Butana (BTN; n = 18), Fogera (FGR; n = 9), Gir (GIR; n = 6), Goffa (GFA; n = 10), Horro (HRO; n = 11), indigenous cattle from Iran (IRN; n = 10), Kenana (KNN; n = 9), Kholmogory (KMG; n = 20), Leiqiong (LQG; n = 3), Limonero (LMN; n = 9) and Ogaden (OGD; n = 17), resulting in a total dataset of 295 domestic cattle samples (Supplementary Table S1). Quality control of raw sequence data, FASTQ files, was first performed using FastQC software (Version 0.4.2) (http://www.bioinformatics.babraham.ac.uk/projects/fastqc), and then sequence trimming and adapter removal were performed by Trimmomatic software (v0.39)⁶². All preprocessed sequencing reads were then mapped to the bovine reference genome (ARS-UCD1.2) using Burrows Wheeler Aligner (BWA, 0.7.18) with default parameters⁶³. The average sequencing coverage was approximately 10.88 X (ranging from 3.53 to 59.17) for each cattle sample (Supplementary Table S1). Prior to variant calling, all cattle BAM files were sorted, and potential PCR duplicates were then removed using command line tools “SortSam” and “MarkDuplicates” implemented in Picard (http://broadinstitute.github.io/picard), respectively. In order to reduce mapping errors, local realignment around INDELs and base quality score calibration (BQSR) were conducted on all BAM files by Genome Analysis Toolkit (GATK) tools⁶⁴. Afterwards, genomic variants were identified and filtered out using GATK tools with the default parameters in all cattle samples (including 295 BAM files). Haplotype phasing and imputing missing genotypes were conducted using the BEAGLE program⁶⁵. All called variants were filtered to be supported by a minimum base quality of 30 (-minQ 30), a minimum mapping quality of 25 (-minMapQ 25) and a minimum genotype quality of 40 (-minGQ 40). In addition, SNPs with more than two alleles and within clusters (> 3 SNPs in a 10-bp window) were filtered out from future analyses⁴⁸.

Population structure, genetic diversity and phylogenetic analysis

FastTree software was used to construct a ML phylogenetic tree from the alignments of nucleotide sequences⁶⁶. To reduce the overall pairwise linkage disequilibrium (LD), whole genome pruning was performed using PLINK software (--indep-pairwise 50 10 0.1)⁶⁷. We applied genome-wide complex trait analysis (GCTA) to compute eigenvectors for PCA⁶⁸. We further used the Bayesian clustering approach in the ADMIXTURE program, with ancestral clusters ranging from 2 to 5 and 10,000 iterations for each run, to identify the degree of Admixture between different populations⁶⁹. Additionally, we used PLINK to calculate the π in windows of 50,000 base pairs (50 kb) for each cattle population⁶⁷. In order to determine the correlations between the θπ values of the populations under investigation, we computed the θπ between different groups (using a 50-kb non-overlapping window) and visualized the results using a scatter plot in the R environment (https://www.r-project.org/).

Selective sweeps and gene set enrichment analyses

To discover genomic regions under selection, two independent statistical methods were applied. The genome-wide weighted FST⁷⁰ was calculated by VCFtools⁷¹, as it is less biased when using populations with unequal samples. The threshold of FST values was set to 0.410 (top 1% of empirical distribution) to determine outliers between studied groups. Furthermore, we calculated θπ along the entire genome using VCFtools⁷¹. Sliding window analyses were carried out for the complete genome with a window size of 50 kb and a step size of 25 kb. By using the Ensemble Variant Effect Predictor (VEP) algorithm (http://asia.ensembl.org/info/docs/tools/vep/index.html), all protein-coding genes discovered by the aforementioned approaches were annotated in the selected regions. Finally, to identify classes of genes and biological categories enriched in the gene lists, gene set enrichment analysis (GSEA) and Kyoto Encyclopedia of Genes and Genomes (KEGG) were conducted by using the ‘g: Profiler’ enrichment analysis tool (https://biit.cs.ut.ee/gprofiler/). The obtained P values were further adjusted by the Benjamini-Hochberg false-discovery rate (BH-FDR).

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization, R.T., J.W. and H.A.N.; writing—original draft preparation, H.A.N., J.T., M.Z., H.L., X.W.; writing—review and editing, R.T., H.A.N.; funding acquisition, R.T., J.W.; All authors have read and agreed to the published version of the manuscript.

Funding

Funding: National Key Research and Development Program of China (2024YFD130010402), Inner Mongolia Agriculture and Animal Husbandry Innovation Fund (2023 CXJJM01, 2025CXJJM09), Foreign young talents program (QN2023005001L), Inner Mongolia Science and Technology Program (2023YFDZ0035).

Data availability

The datasets analysed during the current study are available at NCBI SRA Database (https://trace.ncbi.nlm.nih.gov/Traces/sra) with accession codes: PRJNA256210, PRJEB18113, PRJNA474946, PRJNA431934, PRJNA343262, PRJNA324822, PRJNA494431, PRJNA210523, PRJNA210519, PRJEB28191, PRJNA574857 and PRJNA312138.

Declarations

Competing interests

The authors declare no competing interests.

Institutional review board statement

Not applicable, not empirical research.

Informed consent

Not applicable.