The important economic traits of ruminants result from interactions between genetic background and environmental factors, but key traits such as reproductive performance, feed efficiency, disease resistance, and livestock product quality are often not fully explained by DNA sequence variations alone. Increasing evidence suggests that epigenetic regulation serves as a crucial molecular bridge connecting environmental stimuli with changes in gene expression, allowing organisms to exhibit stable and plastic phenotypic differences without altering the DNA sequence. This review provides a structured synthesis of recent research in the field of epigenetics in ruminants, elucidating how multiple layers of epigenetic mechanisms, including DNA methylation, histone modifications, non-coding RNAs, and higher-order chromatin structures, coordinate to regulate growth, development, reproductive performance, metabolic and immune homeostasis, and livestock product traits across different tissues and developmental stages. These epigenetic marks not only demonstrate high responsiveness to nutrition, management, and environmental stressors, but can exhibit context-dependent stability within the same tissue and physiological stage when environmental conditions are comparable, thereby contributing to the regulation of phenotypic plasticity and offering potential value as predictive biomarkers. Furthermore, epigenetic information can supplement our understanding of phenotypic variation in ways that traditional genomic selection methods are unable to capture, offering new data dimensions for the prediction and improvement of low heritability, environmentally sensitive traits. Overall, integrating epigenetic information with genomic selection strategies may improve the accuracy of ruminant trait prediction and enhance environmental adaptability. This integration also offers a conceptual basis and technical pathway for developing more precise and sustainable breeding systems.

Multi-omics studies have multiplied associations, but many still lack causal resolution and a clear path to application. We present a practical roadmap built on four sequential steps: first, identify signals from genome-wide association studies; second, confirm these signals through regulatory colocalization and transcriptome-wide association analyses; third, integrate the evidence using network analyses and causal inference; and, fourth, test shortlisted candidates through functional and phenotypic validation. The roadmap is supported by three safeguards that make results reliable and reusable: containerized workflows that ensure end-to-end reproducibility, harmonization across batches with concise minimum-information records, and consistent identifier mapping with quality control across data layers. Across four classes of traits—growth and development, carcass and meat quality, reproduction, and environmental adaptation and resilience—we prioritize signals that remain robust across ancestries and environments, highlight modules with explicit regulatory support, and advance candidates that have already progressed to functional testing. Two application tracks follow from this process: integrating stable candidates into selection indices with context-dependent weighting, and recording and targeting mechanistic nodes for nutritional and management interventions. Taken together, this roadmap improves causal interpretability, strengthens cross-population robustness, and shortens the path from statistical association to genetic evaluation and industry uptake.

As essential branches of systems biology, metabolomics and lipidomics systematically reveal the composition, dynamic changes, and biological functions of small-molecule metabolites and lipids using high-throughput analytical techniques. This review examines the application of these omics technologies in evaluating livestock and poultry meat and egg quality, focusing on their roles in elucidating the molecular mechanisms behind key traits such as flavor, tenderness, and nutritional value. By identifying key metabolic markers—including glutamic acid, inosine monophosphate, and specific triglycerides—the intrinsic links between these markers and intramuscular fat deposition, flavor precursor formation, and antioxidant capacity are highlighted. Furthermore, this paper emphasizes the transformative impact of integrating multi-omics data with artificial intelligence (AI). AI-driven analytical frameworks are overcoming the limitations of traditional high-dimensional data processing, enabling robust biomarker discovery, predictive modeling for product quality, and reverse design for genetic improvement. Ultimately, the synergistic application of metabolomics, lipidomics, and AI will drive the development of modern animal husbandry toward intelligent, predictable, and precision-based production.

Background: The Chungtien schizothoracin (Ptychobarbus chungtienensis) is a threatened freshwater fish endemic to the Qinghai–Tibet Plateau and adjacent high-altitude regions of northwestern Yunnan, China. Although a complete mitochondrial genome of P. chungtienensis has been previously reported, direct comparison with a mitogenome generated using high-accuracy long-read sequencing can provide additional information on mitochondrial genome structure and sequence variation. This study aimed to assemble and annotate a complete mitogenome of P. chungtienensis using PacBio HiFi sequencing and to compare its mitogenomic characteristics with previously published Ptychobarbus mitogenomes. Methods: High-molecular-weight genomic DNA from a single specimen was sequenced using PacBio HiFi long-read technology. The mitochondrial genome was assembled using MitoHiFi, annotated using MitoFinder followed by manual curation, and compared with previously published Ptychobarbus mitogenomes. Phylogenetic relationships were evaluated using maximum-likelihood analysis with expanded taxon sampling, and selection pressure on the 13 mitochondrial protein-coding genes was assessed using dN/dS-based branch and branch-site models. Results: The assembled mitogenome is 16,583 bp in length and contains the typical 37 mitochondrial genes, including 13 protein-coding genes, 22 tRNA genes, and 2 rRNA genes, together with a control region and the origin of light-strand replication (O_L). The overall A + T content was 54.97%. Direct comparison with the previously reported 16,970 bp mitogenome showed that the 387 bp length difference was concentrated in non-coding regions, particularly the control region and the tRNA-Thr–tRNA-Pro intergenic region. Phylogenetic analysis based on 22 complete mitogenomes placed the newly assembled P. chungtienensis sequence in a strongly supported mitochondrial clade with Schizothorax macropogon (bootstrap = 100%), whereas the previously reported P. chungtienensis sequence clustered with P. kaznakovi (bootstrap = 100%), indicating that the two P. chungtienensis records represent distinct mitochondrial lineages. The dN/dS values of all 13 mitochondrial protein-coding genes were below 1, and neither branch nor branch–site analyses detected significant evidence of lineage-specific positive selection. Conclusions: This long-read-based mitogenome provides a high-quality genomic resource for P. chungtienensis and reveals substantial mitochondrial sequence and lineage variation among available records. These results provide a basis for comparative mitogenomic and conservation genetic studies while also indicating that species-level phylogenetic relationships and high-altitude adaptation should not be inferred from mitochondrial data alone.