From candidate genes to whole-genome approaches: current insights into the genetic architecture of ovine milk production and quality traits

Abstract

Milk production and composition traits are important characteristics in dairy sheep, influencing milk quality, cheesemaking performance, and the economic value of dairy products. Early genetic studies focused on candidate genes, particularly milk protein genes, endocrine regulators, and lipid-metabolism genes. Among these, the casein gene cluster and the alpha-lactalbumin gene (LALBA) remain consistently replicated loci associated with milk composition traits across breeds and analytical frameworks. The development of genome-wide technologies has expanded our understanding of the genetic architecture underlying ovine dairy traits. Quantitative trait loci (QTL) mapping, genome-wide association studies (GWAS), regional heritability mapping, and single-step GWAS have identified loci associated with milk yield, milk composition, somatic cell score, coagulation properties, and cheesemaking performance. Whole-genome sequencing and selection-signature approaches have expanded the catalogue of candidate genomic regions and biological pathways relevant to dairy performance. These studies indicate that ovine milk traits are polygenic and involve biological pathways related to milk protein synthesis, lipid metabolism, mammary gland development, immune function, tissue remodeling, and environmental adaptation. Recent advances in functional genomics, transcriptomics, and multi-omics integration provide opportunities to prioritize candidate genes and investigate the biological mechanisms underlying dairy phenotypes. Developments in sequence-based genomic resources, structural-variant discovery, and functional genomics are expected to improve the detection of causal variants and support efficient and sustainable breeding strategies. This review summarizes knowledge on candidate genes and genomic regions associated with ovine milk production and quality traits, highlighting the transition from classical candidate-gene studies to genome-wide and post-genomic approaches.

1. Introduction

Sheep milk represents an important resource for dairy production, particularly in the Mediterranean basin and in semi-arid, mountainous and marginal areas (Pulina et al., 2018; Scintu & Piredda, 2007), where local breeds are closely linked to traditional farming systems, biodiversity conservation and the production of high-value dairy products (Albenzio et al., 2016; Balthazar et al., 2017; Park et al., 2007). Although sheep are reared for several economically important products, including wool, meat and milk, their milk is only rarely consumed as fluid milk and is used predominantly for cheesemaking and the production of fermented dairy products because of its high fat, protein, calcium and casein contents and its excellent technological properties (Barlowska et al., 2011; Park et al., 2007; Tamime et al., 2011).

Dairy sheep farming has a long tradition in the Mediterranean basin, southern, central and eastern Europe and the Near East, where it represents an important component of rural economies and traditional livestock systems (Morand-Fehr et al., 2007; Raynal-Ljutovac et al., 2008; Scintu & Piredda, 2007). In recent years, increasing attention has been paid to small-ruminant dairy products, driven by growing interest in local breeds, sustainable production systems and consumer demand for traditional, high-quality, and safe foods (Albenzio et al., 2016; De Devitiis et al., 2023; Selvaggi et al., 2014; Vargas-Bello-Pérez et al., 2022). Many sheep dairy products in the European Union are marketed under geographical indications, such as Protected Designation of Origin (PDO) and Protected Geographical Indication (PGI), which are strongly associated with origin, authenticity, sensory quality, cultural heritage, and sustainability (Albenzio et al., 2016; De Devitiis et al., 2023).

Milk yield, composition, and technological traits are complex phenotypes influenced by both environmental and genetic factors. Breed, parity, stage of lactation, feeding system, season, management practices, and environmental conditions all contribute substantially to phenotypic variability (Correddu et al., 2021; Gaspa et al., 2022; Nguyen, 2022; Raynal-Ljutovac et al., 2008). At the same time, moderate-to-high heritability estimates reported for several dairy traits indicate an important genetic component underlying milk production and its composition (Carta et al., 2023; Correddu et al., 2019; Haile et al., 2019; Selvaggi et al., 2017; Sharma et al., 2015). Milk production is regulated by a complex network of genes involved in mammary gland development, milk protein synthesis, lipid metabolism, lactose synthesis, immune function and energy partitioning, highlighting the multifactorial nature of dairy performance in sheep (Bionaz & Loor, 2008; Moioli et al., 2007; Selvaggi et al., 2017; Sharma et al., 2015).

The genetic dissection of milk production traits in livestock initially relied on candidate gene studies and quantitative trait locus (QTL) mapping approaches. Beginning in the mid-1990s, linkage-based QTL analyses using microsatellite markers were widely employed to identify chromosomal regions affecting complex production traits, including milk yield, milk composition, and dairy performance (Georges et al., 1995; Gutiérrez-Gil et al., 2009; Mateescu & Thonney, 2010). These studies provided the first evidence for the genetic architecture underlying dairy traits and simplified the methodology for identifying major-effect loci. However, the relatively low density of microsatellite-based linkage maps often resulted in broad confidence intervals spanning several megabases, thereby limiting the resolution of QTL localization and the identification of causal genes and mutations (Goddard & Hayes, 2009; Hayes & Goddard, 2001).

The development of high-density SNP arrays represented a major advance in livestock genomics, enabling genome-wide association studies (GWAS) with substantially improved mapping resolution (Hayes & Goddard, 2010; Sharma et al., 2015; VanRaden, 2008). GWAS facilitated the identification of genomic regions associated with economically important traits and revealed that milk production and quality are controlled by numerous loci of small to moderate effect rather than by a limited number of major genes (Hayes & Goddard, 2010; Sharma et al., 2015). More recently, advances in genomic technologies and analytical methodologies, including high-density SNP genotyping, whole-genome sequencing (WGS), single-step genome-wide association studies (ssGWAS), regional heritability mapping, and post-GWAS functional analyses, have substantially improved our understanding of the genetic architecture underlying ovine milk traits (Gebreselassie et al., 2019; Li et al., 2023a; Mastrangelo et al., 2026; Mohammadi et al., 2022; Sutera et al., 2021a). These studies highlight the highly polygenic nature of traits such as milk production, composition and technological properties, influenced by numerous loci involved not only in milk protein synthesis and lipid metabolism but also in mammary gland development, immune function, tissue remodeling and adaptation-related processes. The evolution of genomic approaches and the corresponding expansion of biological knowledge in dairy sheep are summarized in Fig. 1. Despite the considerable progress achieved over the last two decades, the available evidence remains fragmented across different experimental approaches, populations, and analytical frameworks, making a comprehensive synthesis timely to provide an integrated view of the current understanding of the genetic architecture underlying ovine dairy traits. For this reason, a narrative review was considered the most appropriate approach to critically summarize and integrate evidence generated by different experimental designs and genomic methodologies.

Fig. 1.

Evolution of genomic approaches used to investigate ovine milk production and quality traits.

This review provides a comprehensive overview of the current knowledge on candidate genes and genomic regions associated with ovine milk production, composition and technological traits. Particular emphasis is placed on the transition from classical candidate-gene studies to genome-wide and post-genomic approaches, highlighting emerging opportunities offered by functional genomics, multi-omics integration and next-generation genomic resources for dairy sheep breeding.

2. Literature search strategy

Relevant literature was identified through comprehensive searches of the PubMed, Scopus, and Web of Science databases. Searches were last updated on 30 June 2026 using combinations of keywords related to dairy sheep genetics, including ovine, sheep milk, milk production, milk quality, candidate genes, quantitative trait loci (QTL), genome-wide association studies (GWAS), whole-genome sequencing, genomic selection, selection signatures, genetic architecture, and milk traits. No restriction was applied regarding publication year.

Peer-reviewed articles published in English and investigating the genetic basis of milk yield, milk composition, technological properties, and functional traits in dairy sheep were primarily considered eligible. Studies not directly relevant to the genetic or genomic basis of ovine dairy traits, non-English publications, editorials, and purely qualitative studies were excluded. Studies conducted in other livestock species were considered only when relevant to provide methodological or biological context and were not used as primary evidence for ovine-specific genetic associations. Seminal candidate-gene and QTL studies were included to provide historical background, whereas particular emphasis was placed on recent genome-wide association studies, whole-genome sequencing analyses, post-GWAS investigations, and multi-omics approaches. Review articles were primarily used to identify additional relevant references, whereas original research articles constituted the main source of evidence. Additional publications were identified through cross-referencing the bibliographies of selected articles and recent reviews.

Studies were selected based on their relevance to the scope of this narrative review. The aim was to provide a comprehensive and balanced overview of the evolution of genomic approaches applied to ovine dairy traits, rather than conducting a formal systematic review or quantitative evidence synthesis.

3. Candidate genes for milk production and quality traits

Because sheep milk is primarily used for cheese production, milk composition traits, particularly fat and protein contents, are as important as total milk yield. Early genetic studies, therefore, focused on genes with a well-established biological role in milk protein synthesis, lipid metabolism, endocrine regulation of lactation and mammary gland function (Moioli et al., 2007; Ramos et al., 2009) (Table 1). Although genome-wide approaches have greatly expanded the number of loci associated with dairy traits, several of these classical candidate genes remain among the most consistently replicated genomic regions associated with milk production and composition in sheep. The main candidate genes and the major findings from genome-wide studies are summarized in Table 1, Table 2, respectively.

Table 1.

Major candidate genes associated with ovine milk production and quality traits.

Gene symbol	Gene name	OAR	Main associated traits	Biological role	Key references
CSN1S1	Alpha-s1-casein	6	Milk yield, fat percentage, total solids	Major casein protein involved in milk protein composition and cheese-making properties	Mroczkowski et al., 2004; Calvo et al., 2013
CSN1S2	Alpha-s2-casein	6	Protein yield, lactose percentage, solids-not-fat	Structural milk protein influencing milk composition	Corral et al., 2013; Giambra et al., 2010
CSN2	Beta-casein	6	Milk yield, fat percentage, protein percentage	Major component of casein micelles	Corral et al., 2010; Al-Amareen & Jawasreh, 2022
CSN3	Kappa-casein	6	Milk composition, coagulation traits	Casein micelle stabilization and cheese-making properties	Yousefi et al., 2013; Gras et al., 2016
LALBA	Alpha-lactalbumin	3	Protein percentage, fat percentage, milk composition	Regulatory subunit of lactose synthase; lactose synthesis and milk secretion	García-Gámez et al., 2012; Sutera et al., 2021a
LGB	Beta-lactoglobulin	3	Fat percentage, protein percentage, milk composition	Major whey protein affecting milk quality	Dario et al., 2008; Jawasreh et al., 2019
PRL	Prolactin	20	Milk yield, fat percentage, protein percentage	Lactogenesis and mammary gland development	Staiger et al., 2010; Padilla et al., 2018
GH	Growth hormone	11	Milk yield	Growth, energy partitioning and lactation	do Rosário Marques et al., 2006
POU1F1	Pituitary-specific transcription factor 1	1	Milk yield, fat percentage	Regulation of GH and PRL expression	Ozmen et al., 2014
DGAT1	Diacylglycerol O-acyltransferase 1	9	Milk fat percentage, fat composition	Triglyceride synthesis	Scatà et al., 2009; Dervishi et al., 2015
FABP3	Fatty acid-binding protein 3	2	Fat percentage, protein percentage, total solids	Intracellular fatty acid transport	Calvo et al., 2004
SLC27A3	Solute carrier family 27 member 3	1	Fat percentage, protein percentage, lactose content	Fatty acid transport	Kowalewska-Łuczak et al., 2017
FASN	Fatty acid synthase	11	Fatty acid composition	De novo fatty acid synthesis	Crisà et al., 2010; Cesarani et al., 2019
ACACA	Acetyl-CoA carboxylase alpha	11	Fat percentage, fatty acid profile	Fatty acid synthesis	Crisà et al., 2010
SCD	Stearoyl-CoA desaturase	22	Fatty acid composition	Fatty acid desaturation	Cesarani et al., 2019; Crisà et al., 2010

Table 2.

Major genome-wide approaches and key findings for ovine milk production and quality traits.

Analytical framework	Population / Breed	Main genomic regions	Representative candidate genes	Main traits	References
Classical QTL mapping	Churra; East Friesian × Dorset	OAR2, OAR3, OAR12, OAR18, OAR20, OAR23, OAR24	Broad QTL regions	Milk yield, fat percentage, protein percentage	Gutiérrez-Gil et al., 2009; Mateescu & Thonney, 2010
High-density SNP-chip GWAS	Churra	OAR3	LALBA	Protein percentage, fat percentage	García-Gámez et al., 2012
Combined linkage and LD mapping	Sarda	OAR6, OAR10	Casein cluster	Milk, fat and protein yields	Usai et al., 2014
Repeated-measures GWAS	Valle del Belice	OAR1, OAR6, OAR7, OAR10, OAR21, OAR25, OAR26	TTC7B, SUCNR1, DCPS, TENM3, NOB1	Milk production traits	Sutera et al., 2019
Regional Heritability Mapping (RHM)	Valle del Belice	OAR2, OAR3, OAR20	LALBA, aquaporin-related loci	Fat and protein percentage	Sutera et al., 2021a
Weighted single-step GWAS	Valle del Belice	Multiple OARs	Novel production loci	Milk yield, fat yield, protein yield, SCS	Mohammadi et al., 2022
Multi-model GWAS	Crossbred dairy sheep	Multiple OARs	ITPR2, SLC27A6	Milk yield, fat yield, protein yield	Li et al., 2020
GWAS + post-GWAS prioritization	Assaf; Churra	Multiple OARs	71 prioritized genes	Milk production, coagulation traits, cheese yield, SCS	Marina et al., 2021
Dairy processability GWAS	New Zealand dairy sheep	OAR2, OAR3, OAR6, OAR16, OAR18, OAR20, OAR25, OAR26	55 candidate genes	Milk composition, coagulation properties, protein profile	Marshall et al., 2025
Selection-signature analyses	Altamurana	Multiple OARs	PALMD, RFP145	Divergent milk-yield phenotypes	Moioli et al., 2013
Whole-genome resequencing	Dairy vs non-dairy breeds	Genome-wide	FCGR3A, CTSK, CTSS, GHR, SLC29A4, ROR1, TNRC18	Milk production and dairy adaptation	Li et al., 2023a
Low-coverage WGS + ssGWAS	Dairy sheep	Genome-wide	CCSER1, FGGY, HOOK1, NDUFA10, ZNF385D, NWD1	Milk yield and metabolic efficiency	Li et al., 2025
Selection signatures (FST, XP-EHH)	Dairy vs non-dairy breeds	Genome-wide	DHRS3, TNFRSF1B, AADACL4, ARHGEF11, LRRC71	Milk-related adaptive traits	Ebrahimi et al., 2025
Comparative sheep-goat genomics	Sheep and goats	Genome-wide	CLASP1, PDS5B, ZNF831, CCDC73	Dairy performance and adaptation	Akhatayeva et al., 2026

3.1. Milk protein genes: caseins and whey proteins

The most extensively investigated candidate genes are those encoding the major milk proteins. Caseins represent approximately 76–83% of total sheep milk proteins, whereas whey proteins account for about 17–22% (Park et al., 2007). Genetic variants of major milk proteins can influence milk composition, coagulation properties and cheesemaking performance (Amigo et al., 2000). The ovine casein genes CSN1S1, CSN1S2, CSN2, and CSN3 encode alpha-s1-, alpha-s2-, beta- and kappa-casein, respectively, and are located within the casein cluster on ovine chromosome 6 (Árnyasi et al., 2009; Diez-Tascón et al., 2001).

For CSN1S1, several alleles and protein variants have been identified through electrophoretic, chromatographic, and molecular approaches, and associations with milk production traits have been reported in Polish Merino, East Friesian, Assaf, and other dairy sheep breeds (Calvo et al., 2013; Mroczkowski et al., 2004; Selvaggi et al., 2014). Some studies reported significant effects on milk yield, fat percentage, and total solids content, whereas others detected little or no association with milk production or composition traits (Mroczkowski et al., 2004; Calvo et al., 2013). These contrasting results suggest that the effects of CSN1S1 may depend on breed-specific genetic backgrounds, allele frequencies, sample size, lactation stage, and analytical methodology (Calvo et al., 2013; Selvaggi et al., 2014). The CSN1S2 gene encodes αs2-casein and it has received comparatively less attention than other members of the casein cluster. In Dutch East Friesian sheep, CSN1S2 variants were associated with protein yield, whereas specific polymorphisms identified in Merino sheep were associated with non-fat solids and lactose percentage (Corral et al., 2013; Giambra et al., 2010). Although these findings support a potential role of CSN1S2 in milk composition, the number of available studies remains limited and further validation across populations is required (Corral et al., 2013). The CSN2 gene, encoding β-casein, is particularly relevant because β-casein represents one of the major protein fractions of ovine milk and contributes to casein micelle formation and stability (Amigo et al., 2000; Provot et al., 1995). Several CSN2 variants and polymorphisms have been described in sheep populations, and associations with milk yield or fat and protein percentages have been reported in Awassi, Merino, and other dairy breeds (Al-Amareen & Jawasreh, 2022; Corral et al., 2010). However, studies conducted in East Friesian and Lacaune sheep have not always confirmed these associations (Giambra et al., 2014). Overall, the available evidence suggests that CSN2 contributes to variation in milk traits, although the magnitude and direction of its effects appear to vary among breeds. The CSN3 gene, encoding κ-casein, has been less consistently associated with ovine milk production traits. Positive associations with lactose percentage and milk composition have been reported in Zel sheep and other local populations (Yousefi et al., 2013). In contrast, studies conducted in East Friesian, Teleorman Black Head, and Awassi sheep generally failed to detect significant effects on milk yield or major milk components (Gras et al., 2016; Staiger et al., 2010). These findings suggest that CSN3 may influence specific compositional or technological characteristics of milk. However, its contribution to quantitative production traits appears less pronounced than that reported for other members of the casein cluster.

Among whey protein genes, LALBA is one of the most extensively validated candidates associated with ovine milk composition traits. LALBA encodes α-lactalbumin, the regulatory subunit of the lactose synthase complex, and therefore plays a central role in lactose synthesis, milk volume regulation and osmotic milk secretion during lactation (Moioli et al., 2007; Park et al., 2007). Beyond its direct role in milk synthesis, increasing attention has recently been devoted to the genetic architecture of lactose content itself. Studies in dairy sheep have reported moderate heritability estimates for milk lactose concentration together with significant genetic correlations with milk yield and udder-health traits, suggesting that lactose content may represent an informative indicator of mammary gland functionality and lactation efficiency (Carta et al., 2023; Mastrangelo et al., 2026). In dairy sheep, a major genomic region on OAR3 harboring the LALBA locus has been repeatedly captured across diverse populations and analytical frameworks; this includes initial QTL analyses and subsequent high-density GWAS in Churra sheep, which identified a functional non-synonymous variant (p.Val27Ala) affecting milk fractions, as well as regional heritability mapping (RHM) in the Valle del Belice breed (García-Gámez et al., 2012; Gutiérrez-Gil et al., 2009; Sutera et al., 2021a). Compared with many other candidate genes whose effects appear breed-specific or inconsistently replicated, the convergence of evidence from QTL mapping, GWAS, WGS, and regional heritability mapping supports LALBA as one of the most consistently replicated loci associated with milk composition traits in dairy sheep (García-Gámez et al., 2012; Marina et al., 2020a).

The LGB gene, encoding β-lactoglobulin, has also been extensively investigated because of its role as the major whey protein in ruminant milk and its potential influence on milk composition and processing properties (Amigo et al., 2000; Kawecka & Radko, 2011; Mastrangelo et al., 2012). Several studies have reported associations between LGB polymorphisms and milk fat content, protein content, or other compositional traits in Awassi, Merino, East Friesian, Zel, and Italian local breeds (Dario et al., 2008; Jawasreh et al., 2019; Triantaphyllopoulos et al., 2017; Yousefi et al., 2013). However, other studies failed to detect significant associations with milk yield or major milk components (Kawecka & Radko, 2011; Staiger et al., 2010). Overall, the available evidence suggests that LGB may contribute to variation in milk composition and technological properties, but its effects appear less consistent and more breed-dependent than those reported for the casein cluster or the LALBA region (Amigo et al., 2000; Selvaggi et al., 2014).

3.2. Genes involved in lactation regulation and endocrine pathways

Owing to the central role of prolactin in mammary gland development, lactogenesis, sustained milk secretion, and milk protein gene regulation, the PRL gene has been widely regarded as a biologically plausible candidate influencing dairy traits (Bionaz & Loor, 2008; Ramos et al., 2009). Associations between PRL polymorphisms and milk yield, fat content, and protein percentage have been reported in several dairy sheep breeds, including Serra da Estrela, East Friesian, and Spanish Merino sheep (do Rosário Marques et al., 2006; Padilla et al., 2018; Staiger et al., 2010). However, favorable genotypes are not consistent across populations, suggesting that the effects of PRL may depend on genetic background, breed structure, and environmental conditions (Padilla et al., 2018; Staiger et al., 2010).

The growth hormone axis has also received considerable attention for its roles in growth, energy partitioning and lactation physiology. The GH gene has been associated with milk yield in Serra da Estrela sheep (do Rosário Marques et al., 2006), whereas POU1F1, a pituitary transcription factor regulating both growth hormone and prolactin expression, has been associated with milk yield and fat content in Sakiz sheep (Ozmen et al., 2014). Although less extensively investigated in dairy sheep than in cattle, GHR remains a relevant functional candidate because of its central role in growth hormone signalling and nutrient allocation during lactation (Hayes & Goddard, 2010; Sharma et al., 2015). Recent genome-wide and selection-signature studies have further highlighted GHR and related endocrine and metabolic pathways as biological processes that may contribute to variation in dairy performance (Akhatayeva et al., 2026; Li et al., 2023a). Overall, studies on endocrine-related genes support the importance of hormonal regulation in shaping dairy performance. Nevertheless, compared with milk protein genes, the evidence for individual endocrine candidates remains less consistent, reflecting the complexity of lactation physiology and the likely contribution of multiple interacting pathways rather than major individual loci.

3.3. Genes involved in lipid metabolism and milk fat traits

Milk fat traits have been extensively investigated through genes involved in triglyceride synthesis, fatty acid transport, and lipid metabolism because milk fat is a major determinant of both the nutritional and technological value of sheep milk (Park et al., 2007; Raynal-Ljutovac et al., 2008). Among these genes, DGAT1 encodes diacylglycerol O-acyltransferase 1, the enzyme catalysing the final step of triglyceride synthesis (Chen & Farese, 2000; Coleman & Bell, 1976). Several DGAT1 polymorphisms have been identified in sheep, including variants located in untranslated, intronic, and coding regions (Dervishi et al., 2015; Scatà et al., 2009). Associations between DGAT1 variants and milk fat content have been reported in Sarda, Altamurana, and Gentile di Puglia sheep (Moioli et al., 2013; Scatà et al., 2009), whereas studies in Assaf, Turcana, and other breeds have yielded weaker or inconsistent results (Dervishi et al., 2015; Tăbăran et al., 2014). Consequently, although DGAT1 remains a biologically plausible candidate for milk fat traits, its effects appear less consistent in sheep than in dairy cattle, where DGAT1 represents one of the major genes affecting milk fat yield and composition (Hayes & Goddard, 2010).

Additional lipid-related candidate genes include FABP3, SLC27A3, and SLC27A6, which are involved in fatty acid transport and cellular lipid metabolism. Polymorphisms in FABP3 have been associated with milk fat content in Manchega sheep and with fat, protein, and total solids content in Slovak breeds (Calvo et al., 2004; Kowalewska-Łuczak et al., 2017). Variants in SLC27A3 have been associated with fat, protein, lactose, and urea contents (Kowalewska-Łuczak et al., 2017), whereas SLC27A6 has emerged as a promising candidate gene for milk yield and lactation traits in GWAS analyses of dairy sheep populations (Li et al., 2020). Together these findings suggest that genes involved in fatty acid uptake and intracellular transport contribute to variation in both milk composition and production traits.

Beyond total fat percentage, increasing attention has been paid to the genetic architecture of milk fatty acid composition because the relative abundance of saturated, monounsaturated and polyunsaturated fatty acids, as well as the proportion of odd- and branched-chain fatty acids, influences both the nutritional value and processing properties of dairy products (Nudda et al., 2021; Park et al., 2007; Vargas-Bello-Pérez et al., 2022). In this context, genes involved in de novo fatty acid synthesis and desaturation, including FASN, ACACA, and SCD, have attracted considerable interest (Crisà et al., 2010; Moioli et al., 2007). Variants within these genes have been associated with differences in milk fatty acid profiles and with traits related to milk fat quality in several sheep populations (Crisà et al., 2010). Although the available evidence remains more limited than in cattle, fatty acid profile represents a promising target for breeding programs aimed at improving the nutritional quality of sheep milk and dairy products (Cesarani et al., 2019; Sharma et al., 2015). The increasing availability of detailed milk fatty acid phenotypes, together with genomic, transcriptomic, and other post-genomic resources, is expected to clarify further the contribution of lipid-metabolism pathways to milk quality traits in dairy sheep (Hasin et al., 2017; Vargas-Bello-Pérez et al., 2022).

Collectively, the available evidence indicates that although several candidate genes have been repeatedly associated with ovine milk production and quality traits, many reported effects remain breed-specific and explain only a limited proportion of the observed phenotypic variation. The inconsistent associations reported across breeds likely reflect a combination of biological and methodological factors. Differences in allele frequencies can affect the ability to detect associations, particularly when alternative alleles occur at low frequency or are nearly fixed within a population. Population structure, differences in breed genetic architecture, sample size and statistical power, phenotype definition, environmental conditions, and analytical models may further contribute to heterogeneous findings among studies (Sharma et al., 2015). In addition, the effect of a given variant may depend on the surrounding genetic background, including differences in linkage disequilibrium with the actual causal variant and potential interactions with other loci. Consequently, associations detected for genes such as CSN1S1, LGB, and DGAT1 in one breed may not necessarily be replicated with the same magnitude or direction in another. These inconsistencies emphasize the importance of validating candidate-gene associations across independent populations and further highlight the inherently polygenic architecture of dairy traits. This complexity has contributed to the progressive transition from hypothesis-driven candidate-gene studies toward genome-wide approaches capable of capturing a broader spectrum of genetic variation underlying ovine milk production and quality traits.

4. QTL mapping and SNP-chip GWAS

The genetic dissection of ovine dairy traits initially relied on family-based linkage mapping approaches. These studies used low-density microsatellite markers to identify the first QTL associated with milk yield and composition traits (Gutiérrez-Gil et al., 2009; Mateescu & Thonney, 2010). Despite their contribution to the understanding of the genetic basis of dairy performance, linkage-mapping approaches were constrained by limited marker density and low recombination resolution, often resulting in broad confidence intervals spanning several megabases (Gutiérrez-Gil et al., 2009; Hayes & Goddard, 2010). Several QTL regions were identified on Ovis aries chromosomes (OARs) 2, 3, 12, 18, 20, 23, and 24, although identifying causal variants remained challenging (Gutiérrez-Gil et al., 2009; Mateescu & Thonney, 2010). Although these studies provided the first genomic framework for ovine dairy traits, the broad confidence intervals associated with linkage mapping limited the identification of causal genes and reduced their direct applicability to breeding programmes.

The development of high-density single-nucleotide polymorphism (SNP) arrays represented a major advance in livestock genomics, enabling GWAS with substantially improved mapping resolution and statistical power (Hayes & Goddard, 2010; Sharma et al., 2015). In Churra sheep, GWAS refined a major QTL for protein and fat percentage to a narrow genomic interval encompassing the third intron of the LALBA gene on OAR3, providing strong support for previous linkage-based findings (García-Gámez et al., 2012). In Sarda sheep, combined linkage and linkage disequilibrium analyses identified a major signal for protein content on OAR6, corresponding to the casein gene cluster, together with an additional region on OAR10 associated with milk, fat and protein yields (Usai et al., 2014).

Subsequent genome-wide studies further expanded the catalogue of candidate loci associated with dairy traits. Repeated-measures GWAS identified significant markers across different chromosomes (OARs 1, 6, 7, 10, 21, 25, and 26) highlighting TTC7B, SUCNR1, DCPS, TENM3, and NOB1 as promising candidate genes for milk production traits (Sutera et al., 2019). These findings were subsequently supported by RHM, which confirmed the importance of the LALBA region and nearby aquaporin-related loci on OAR3, as well as genomic regions on OAR2 and OAR20 associated with fat percentage (Sutera et al., 2021a). More recently, Weighted single-step GWAS (WssGWAS) identified additional genomic regions associated with milk, fat and protein yields and somatic cell score (SCS), extending genomic investigations to functional traits related to milk quality and udder health (Mohammadi et al., 2022). The transition from linkage mapping to SNP-based GWAS represented a major methodological advance, substantially increasing mapping resolution and enabling the identification of numerous novel loci. At the same time, GWAS highlighted the highly polygenic nature of ovine milk traits, with most genomic regions explaining only a small proportion of the overall genetic variance.

Importantly, genome-wide analyses not only confirmed the relevance of several classical candidate genes, such as LALBA and the casein cluster, but also identified numerous previously unsuspected loci involved in mammary development, metabolism, immune function, and tissue remodeling. Although these loci were identified using different analytical approaches and in diverse sheep populations, several chromosomal regions have been consistently replicated across independent studies. The approximate chromosomal distribution of the principal candidate genes and recurrent QTL/GWAS regions is summarized in the conceptual overview presented in Fig. 2.

Fig. 2.

Conceptual overview of the principal genomic regions implicated in ovine milk production, composition, and technological traits. The diagram summarizes the main ovine chromosomes (OAR) harboring repeatedly reported candidate genes, quantitative trait loci (QTL), and genomic regions identified by genome-wide association studies (GWAS). Loci are color-coded according to their primary biological pathways or technological relevance, illustrating the convergence between classical candidate-gene studies and modern genome-wide approaches. The chromosomal locations shown are schematic and are intended solely to illustrate the approximate distribution of the principal genomic regions discussed in this review. The figure is not intended to represent precise physical positions, QTL intervals, or genome coordinates.

5. Emerging genes from genomic studies

Genome-wide studies have considerably expanded the range of candidate genes associated with dairy traits beyond the classical milk protein and lipid-metabolism genes traditionally investigated in sheep. Genome-wide selective sweep analyses have identified genomic regions under selection harboring candidate genes, including PALMD and RFP145, in Mediterranean sheep breeds adapted to marginal environments (Moioli et al., 2013). Similarly, multi-model GWAS and comparative genomic analyses in specialized crossbred and international dairy populations have identified ITPR2 and SLC27A6 (Li et al., 2020), as well as SPATA6 and GAL3ST3 (Jawasreh et al., 2022), as candidate genes associated with milk production traits.

An important outcome of these studies is that many newly identified candidate genes are not part of pathways traditionally associated with lactation. Whereas early candidate-gene studies focused primarily on milk proteins, endocrine regulators and lipid metabolism, genome-wide approaches have increasingly highlighted genes involved in cellular signaling, immune response, tissue remodeling, energy metabolism, and mammary gland development (Li et al., 2020, 2023a; Marina et al., 2021; Marshall et al., 2025; Mohammadi et al., 2022; Sutera et al., 2021b). Several of these genes are involved in cellular signaling, nutrient transport, metabolic regulation, and developmental processes, further supporting the view that milk production and quality traits are influenced by a broad range of biological mechanisms that extend beyond the direct synthesis of milk components.

Similar trends have been observed in other livestock species, including cattle and goats, where genome-wide studies increasingly identify genes involved in regulatory, developmental, and adaptive processes rather than exclusively in structural components of milk synthesis (Akhatayeva et al., 2026; Sharma et al., 2015). Although several of these emerging candidates still require functional validation through expression quantitative trait loci (eQTL) mapping (Shi, 2020), gene expression studies and experimental approaches, their repeated identification across populations and analytical approaches supports an increasingly complex view of the genetic architecture underlying dairy performance in sheep. A notable outcome of genome-wide analyses is the progressive shift from structural milk-protein genes toward regulatory, developmental, and adaptive pathways. This broader spectrum of biological functions suggests that dairy performance is influenced not only by genes directly involved in milk synthesis but also by mechanisms affecting mammary function, metabolic efficiency, immune competence and environmental responsiveness.

5.1. Expanding GWAS to milk processing and cheesemaking traits

Because sheep milk is predominantly utilized for high-value cheese production rather than direct fluid consumption, genomic research has progressively shifted toward technological and processing phenotypes (Gaspa et al., 2016; Marina et al., 2020b, 2021; Marshall et al., 2025). Traits such as milk coagulation properties (MCP, including rennet clotting time and curd firmness), individual cheese yield, and detailed milk protein fractions are of paramount economic relevance because they directly dictate manufacturing efficiency and the sensory profile of the final dairy product (De Devitiis et al., 2023; Raynal-Ljutovac et al., 2008). The application of genome-wide approaches to these phenotypes has revealed that the genetic architecture of milk processability is highly complex and only partially overlaps with loci associated with total milk yield and gross composition.

In dairy sheep populations with an intense cheesemaking tradition, early association mapping identified polymorphisms within the casein gene cluster associated with curd firming rates and cheese yield, independently of total daily milk volume (Noce et al., 2016). To further untangle these specialized pathways, recent studies have combined high-density GWAS with post-GWAS functional prioritization analyses. In Assaf and Churra dairy sheep, this multi-tiered approach prioritized 71 candidate genes associated with one or more of the investigated traits, including milk production, somatic cell score, milk coagulation properties, and individual laboratory cheese yield (Marina et al., 2021).

Similarly, large-scale genomic mapping in Southern Hemisphere dairy sheep populations detected 87 significant SNPs and 55 candidate genes associated with individual milk protein profiles and fine dairy processability parameters (Marshall et al., 2025). These studies indicate that cheesemaking and milk processing traits have a complex polygenic architecture involving multiple genomic loci. Several prioritized regions include genes involved in biological processes related to micellar stability, calcium transport, and proteolytic activity. These findings reinforce the concept that milk processability and cheesemaking aptitude constitute complex traits with a partially distinct genetic basis from conventional production traits. At the same time, they support the inclusion of direct milk processability and technological phenotypes alongside traditional production traits in future genomic selection matrices and breeding indices to preserve the cheesemaking aptitude of specialized dairy sheep breeds (Marina et al., 2021; Marshall et al., 2025).

The partial overlap between genomic regions associated with milk composition and technological traits also raises the question of whether shared genetic signals reflect true pleiotropy or linkage disequilibrium between distinct causal variants. This distinction is particularly relevant for complex genomic regions such as the casein gene cluster on OAR6, where several closely linked genes may influence milk protein composition and coagulation properties. A causal variant affecting milk composition could exert pleiotropic effects on coagulation and cheesemaking traits through changes in milk protein characteristics. Alternatively, associations with multiple traits may result from distinct functional variants that are maintained in linkage disequilibrium within the same genomic region. Current association studies generally do not allow these mechanisms to be clearly distinguished. Higher-resolution fine-mapping, haplotype-based and conditional association analyses, together with functional validation, will therefore be important to determine whether shared signals for milk composition and technological traits reflect true pleiotropy or linked causal variants.

6. Whole-genome sequencing, low-coverage WGS and selection-signature approaches

Whole-genome and sequence-based approaches are increasingly used to overcome some limitations of SNP-chip GWAS. Commercial SNP arrays capture only a subset of common variants and may miss rare variants, structural variants and population-specific polymorphisms (Hayes & Goddard, 2010; Sharma et al., 2015). Whole-genome resequencing and low-coverage WGS, particularly when combined with imputation, provide denser genomic information and improve the detection of variants associated with complex dairy traits (Li et al., 2020, 2025).

Whole-genome resequencing studies have expanded the catalogue of candidate genes potentially involved in ovine milk production and quality. Selection-signature analyses comparing dairy and non-dairy sheep breeds identified genomic regions under selection harboring genes involved in immune response, mammary development, metabolism and lactation biology, including FCGR3A, CTSK, CTSS, ARNT, GHR, SLC29A4, ROR1, and TNRC18 (Li et al., 2023a). Additional whole-genome analyses in East Friesian sheep highlighted genomic regions showing signatures of artificial and natural selection, further supporting the potential contribution of metabolic and developmental pathways to dairy performance (Li et al., 2022, Rezvannejad et al., 2022).

Low-coverage whole-genome sequencing (LcWGS) has recently emerged as a cost-effective alternative for large-scale genomic studies. A recent study combining low-coverage sequencing, imputation, and WssGWAS identified CCSER1, FGGY, and HOOK1 as candidate genes for milk yield, and NDUFA10, ZNF385D, and NWD1 as candidates for milk yield adjusted for metabolic body weight (Li et al., 2025). Beyond the specific candidate genes identified, this study demonstrated the feasibility of integrating low-coverage sequencing, imputation and genome-wide analyses within practical breeding programs for dairy sheep.

However, the performance of LcWGS combined with genotype imputation strongly depends on the size and breed composition of the reference panel and its genetic relationship with the target population. Studies in sheep have shown that imputation accuracy generally improves with increasing reference population size and with greater genetic relatedness between reference and target populations, whereas genetically distant or poorly represented populations may show substantially lower accuracy (Bolormaa et al., 2019; Moghaddar et al., 2015). This limitation is particularly relevant for rare and population-specific variants, which tend to be imputed less accurately (Bolormaa et al., 2019). Recent applications of LcWGS in sheep have further demonstrated that sequencing depth and reference-panel composition are important determinants of imputation performance (Li et al., 2025). In downstream association analyses, imputation errors can reduce statistical power by attenuating true genotype–phenotype associations, while poorly imputed variants or systematic differences in imputation accuracy among populations may also generate spurious associations and increase the risk of false-positive findings. The development of larger and breed-representative reference panels, together with appropriate assessment of imputation accuracy and stringent post-imputation quality control, will therefore be essential for the reliable implementation of LcWGS in dairy sheep genomics.

Recent selection-signature studies have also provided additional insights into the genetic basis of dairy adaptation. Using the Sheep HapMap SNP50 dataset, Ebrahimi et al. (2025) compared dairy and non-dairy breeds through F_ST and XP-EHH analyses and identified genomic regions under selection harboring candidate genes potentially involved in dairy performance, including DHRS3, TNFRSF1B, AADACL4, ARHGEF11, and LRRC71. Similarly, a comparative whole-genome study of sheep and goats identified convergent signatures of selection potentially related to dairy performance, highlighting CLASP1, PDS5B, ZNF831, and CCDC73. Transcriptomic evidence further supported a potential role of CLASP1 in mammary tissue remodeling and lactation-related processes (Akhatayeva et al., 2026).

Although the specific candidate genes identified by WGS and sequencing-based studies differ among populations, breeds and analytical approaches, a consistent biological pattern is emerging. Many of the detected loci are involved in mammary gland development, immune regulation, epithelial remodeling, endocrine signaling, lipid metabolism, energy balance, and adaptation-related processes (Ebrahimi et al., 2025; Li et al., 2022, 2023a). This broader functional spectrum suggests that variation in milk production and quality traits is influenced not only by genes directly involved in milk secretion, but also by pathways affecting animal physiology, metabolic efficiency and the biological response of mammary tissue to environmental and management conditions (Hayes & Goddard, 2010; Marina et al., 2021; Marshall et al., 2025). In general, these findings indicate that the genetic architecture of ovine milk traits extends well beyond the classical milk protein genes and supports a highly polygenic model in which numerous loci of small to moderate effect contribute to dairy performance. Nevertheless, many of these genes remain putative candidates and require validation in independent populations, functional annotation, gene expression analyses and, ideally, causal variant identification (Hasin et al., 2017; Li et al., 2025). Although many sequence-derived candidate genes remain to be functionally validated, sequencing approaches have substantially expanded the spectrum of biological processes implicated in dairy performance. Their main contribution lies not only in the discovery of novel loci but also in revealing the importance of rare variants, adaptive processes, and population-specific genetic diversity that are only partially captured by SNP-array studies.

These methodological advances have progressively increased both the resolution and the biological interpretability of genomic analyses in dairy sheep. From hypothesis-driven candidate-gene studies to sequence-based and integrative multi-omics approaches, each strategy has contributed distinct insights into the genetic architecture of milk production and quality traits. A comparative overview of the principal genomic approaches is presented in Table 3.

Table 3.

Comparison of genomic approaches used to investigate ovine milk production and quality traits.

Approach	Main genomic resource	Main advantages	Main limitations	Representative findings
Candidate gene analysis	Individual genes	Biologically driven; simple interpretation	Biased toward known genes; limited genome coverage	Casein cluster, LALBA, DGAT1
Linkage/QTL mapping	Microsatellite markers	First identification of genomic regions controlling milk traits	Low mapping resolution; broad confidence intervals	QTL on OAR2, OAR3, OAR6, OAR20
SNP-chip GWAS	Medium/high-density SNP arrays	Genome-wide discovery of associated loci	Limited detection of rare and structural variants	Confirmation of LALBA and identification of novel loci
Regional heritability mapping	SNP windows	Detects regional effects not captured by single SNPs	Lower positional resolution	LALBA region, OAR2 and OAR20
ssGWAS / WssGWAS	SNPs + pedigree information	Greater power in breeding populations	Computationally demanding	Milk yield, protein yield, somatic cell score
Whole-genome sequencing	Sequence variants	Detection of rare and potentially causal variants	Higher cost and bioinformatic complexity	Novel candidate genes and population-specific variants
Selection-signature analysis	Genome-wide sequence/SNP data	Identifies adaptive loci shaped by selection	Does not directly demonstrate trait association	Genes involved in mammary development, metabolism and adaptation
Functional genomics / Multi-omics	Transcriptomic, proteomic and metabolomic data	Prioritization of candidate genes and biological pathways	Limited availability of functional datasets	Identification of regulatory mechanisms and causal pathways

7. From association signals to functional genomics

Considerable progress has been achieved through GWAS and sequencing-based approaches, but the identification of associated genomic regions represents only the first step toward understanding the biological mechanisms underlying milk production traits. Increasingly, genomic studies are being integrated with transcriptomic, proteomic, and metabolomic analyses to prioritize candidate genes and identify functional pathways (Hasin et al., 2017; Subramanian et al., 2020; Suravajhala et al., 2016). The Functional Annotation of Animal Genomes (FAANG) initiative is providing standardized epigenomic, transcriptomic and chromatin-accessibility resources that are expected to facilitate the functional interpretation of GWAS loci and the identification of causal regulatory variants across livestock species (Giuffra et al., 2019).

Gene expression studies in mammary tissue have revealed differential expression of genes involved in milk synthesis, immune regulation, and mammary gland development, providing independent support for several loci identified through association analyses (Bionaz & Loor, 2008; Suarez-Vega et al., 2015). The integration of GWAS results with transcriptomic information can also facilitate the identification of regulatory variants and eQTL, helping to distinguish causal genes from linked markers and improving the biological interpretation of association signals (Suravajhala et al., 2016).

Comprehensive population- and tissue-specific eQTL resources remain limited in sheep; however, complementary approaches can help interpret non-coding association signals and prioritize candidate genes. Statistical fine-mapping and linkage disequilibrium information can be integrated with tissue-specific regulatory annotations, including promoters, enhancers, chromatin accessibility, and histone modifications generated by FAANG and related functional genomics resources (Giuffra et al., 2019; Xie et al., 2026). Recent multi-omics annotation of the sheep genome has substantially expanded the catalogue of tissue-specific regulatory elements and provides a useful framework for interpreting the potential regulatory role of non-coding variants (Xie et al., 2026). For dairy traits, regulatory information from mammary tissue can be further integrated with RNA-seq data from the lactating mammary gland to prioritize genes located within or functionally linked to associated genomic regions and showing expression patterns relevant to lactation (Suarez-Vega et al., 2015). The convergence of statistical, regulatory, and transcriptomic evidence can therefore provide stronger support for prioritizing variants and candidate genes for subsequent functional validation. Until larger population- and tissue-specific eQTL resources become available in sheep, these complementary approaches provide a practical framework for linking non-coding GWAS signals to biologically plausible candidate genes.

A substantial proportion of the additive genetic variance underlying milk production and quality traits remains unexplained. This phenomenon, commonly referred to as “missing heritability”, has been widely discussed in both human and livestock genetics and may reflect the contribution of rare variants, structural variation, gene–gene interactions (epistasis), non-additive genetic effects, gene–environment interactions and regulatory mechanisms that are not fully captured by conventional association analyses (Hayes & Goddard, 2010; Manolio et al., 2009; Sharma et al., 2015). Integrative approaches that extend beyond common SNP variation may therefore be required to characterize additional sources of genetic variance.

Structural variants (SVs) may represent an important component of this unexplained variation because conventional SNP arrays are not designed to comprehensively capture large insertions and deletions, copy-number variants, inversions, and other complex genomic rearrangements. This limitation has been demonstrated in sheep, where many sequence-resolved SVs showed limited linkage disequilibrium with surrounding SNPs and were not effectively tagged by probes included in the widely used OvineSNP50 BeadChip (Li et al., 2023b). Structural variants may influence phenotypes by altering gene dosage, disrupting coding sequences, or modifying regulatory elements, and their contribution may be particularly relevant in repetitive, copy-number-variable, or otherwise structurally complex genomic regions. Long-read sequencing and pangenome approaches can improve the detection and representation of such variation by resolving large genomic rearrangements, non-reference sequences, and alternative haplotypes that may be absent or poorly represented in a single linear reference genome (Li et al., 2023b; Qiao et al., 2024). These approaches therefore provide a complementary layer of genomic information to standard SNP-based analyses and may help explain part of the unexplained genetic variance underlying complex dairy phenotypes. However, the specific contribution of SVs to ovine milk production and technological traits remains largely unexplored and will require population-scale association studies and functional validation.

Future advances in ovine dairy genomics will increasingly depend on integrating sequence-based genomic variation, including SNPs and SVs, with transcriptomic, proteomic, metabolomic, and other functional datasets to link genetic variation to biological function and production phenotypes. Such approaches are expected to improve candidate-gene prioritization, facilitate the identification of causal variants, and provide a more comprehensive understanding of the molecular mechanisms underlying milk production, composition, and technological traits in dairy sheep (Hasin et al., 2017; Li et al., 2025). Ultimately, the major challenge facing ovine dairy genomics is no longer the identification of associated loci, but rather the translation of association signals into biological mechanisms. Functional validation and multi-omics integration will therefore be essential to distinguish causal variants from linked markers and to improve the practical exploitation of genomic discoveries in breeding programs.

8. Concluding remarks and future perspectives

Research on the genetics of ovine milk production and quality traits has evolved from the study of a limited number of biologically plausible candidate genes to the investigation of complex polygenic architectures through genome-wide approaches. Among the loci identified to date, the casein gene cluster and the LALBA region remain the most consistently replicated candidates across breeds, populations and analytical frameworks (García-Gámez et al., 2012; Gutiérrez-Gil et al., 2009; Sutera et al., 2021a).

At the same time, GWAS, ssGWAS, WGS, and selection-signature studies have revealed a substantially broader network of genes involved in metabolism, endocrine regulation, immune response, mammary gland development, tissue remodeling, and other biological processes potentially relevant to dairy performance (Li et al., 2023a; Marshall et al., 2025; Mohammadi et al., 2022).

The current body of evidence indicates that milk yield, composition, and technological properties are controlled by numerous loci of small to moderate effect rather than by a limited number of major genes (Marina et al., 2021; Marshall et al., 2025). This increasingly complex picture is consistent with the concept of missing heritability, whereby classical candidate genes explain only a fraction of the genetic variation observed for dairy traits (Hayes & Goddard, 2010; Manolio et al., 2009). Future studies should therefore move beyond the simple identification of associated loci and focus on the validation of candidate variants and the characterization of the biological pathways underlying complex dairy phenotypes.

Integrative approaches combining genomics, transcriptomics, proteomics, metabolomics, and detailed phenotyping of milk composition, fatty acid profile, coagulation properties, and cheesemaking performance are expected to provide a more complete understanding of the biological mechanisms underlying dairy performance and milk quality (Hasin et al., 2017; Marina et al., 2021; Marshall et al., 2025).

Genotype-by-environment interactions represent an additional component of the genetic architecture of dairy traits, particularly when animals are raised under contrasting production systems. Studies in dairy sheep have shown that the expression of genetic differences in milk production can vary across management and production environments, although the magnitude of G × E interactions depends on the populations and environmental conditions considered (Sanna et al., 2002). Consequently, genotypes selected for high production under intensive and relatively favorable conditions may not necessarily maintain the same relative performance under marginal or extensive systems, where animals are exposed to greater environmental constraints. This aspect is particularly relevant in Mediterranean production systems, characterized by high temperatures, seasonal variation in feed availability, and low-input management conditions. Evidence also indicates that selection objectives based predominantly on milk production may involve trade-offs with specific components of environmental resilience. For example, a negative genetic relationship between milk production and heat tolerance has been reported in Mediterranean dairy sheep (Finocchiaro et al., 2005), while genomic analyses have identified heritable variation in resilience to climatic fluctuations and a negative genetic relationship with milk production (Tsartsianidou et al., 2021). These findings suggest that breeding objectives focused primarily on increasing production may not adequately account for performance stability and resilience under marginal or climate-stressed environments. Future breeding strategies should therefore consider production alongside robustness, heat tolerance, and locally relevant adaptive traits, particularly in populations maintained under extensive or semi-extensive Mediterranean conditions.

Advances in whole-genome sequencing, structural variant discovery, and graph-based pangenomes are likely to further improve the identification of causal variants and the characterization of breed-specific genomic diversity (Li et al., 2023b; Liu, 2025). Ultimately, the combination of large multi-breed populations, sequence-based genomic resources, functional genomics approaches, high-resolution phenotyping, and environmental information will contribute to a more comprehensive understanding of the biological basis of dairy traits and support sustainable genetic improvement and biodiversity conservation in sheep breeding systems. These advances provide an increasingly robust foundation for the implementation of precision breeding strategies aimed at improving milk quality and cheesemaking performance while maintaining resilience in dairy sheep production systems.

Ethics declaration

Ethics committee approval was not required. This review relied on published data, no new data and no new samples were collected, and as such no animal ethics approval was required.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

During the preparation of this manuscript, the authors used ChatGPT solely to improve the grammar and readability of the text.

Statement for studies

This article is a review of the published literature and does not contain any studies with human participants or animals performed by the authors. Therefore, ethical approval and informed consent were not required.

Funding sources

This research received no specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

CRediT authorship contribution statement

Salvatore Mastrangelo: Writing – review & editing, Writing – original draft, Validation, Supervision, Conceptualization. Alessia Benanti: Writing – review & editing, Data curation. Serena Tumino: Writing – review & editing, Data curation. Andrea Criscione: Writing – review & editing, Supervision, Data curation. Alberto Cesarani: Writing – review & editing, Validation, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.