Animal Molecular Breeding 2024, Vol.14, No.3, 239-251 http://animalscipublisher.com/index.php/amb 240 This study comprehensively analyzes the progress and applications of genomic selection (GS) in livestock breeding, covering its technical foundation, applications across various livestock species, and future prospects, providing insights into how GS can further revolutionize animal breeding to enhance efficiency, productivity, and sustainability in livestock production. 2 Principles of Genomic Selection 2.1 The genetic basis of livestock traits Genomic selection (GS) is an innovative approach in livestock breeding that leverages the comprehensive analysis of genetic markers across the entire genome to predict an animal's breeding value. This method has revolutionized the field by enabling breeders to make more informed and accurate selection decisions. Unlike traditional methods that focus on observable traits or a limited number of genetic markers, GS utilizes high-density single nucleotide polymorphism (SNP) chips to evaluate thousands of markers simultaneously. This approach allows for the capture of both large and small genetic effects, leading to more precise genetic predictions and faster genetic progress. The principles of GS are rooted in understanding the genetic architecture of traits and applying advanced statistical models to integrate vast amounts of genomic data. As the technology and computational methods continue to evolve, the application of GS is expanding across various livestock species, enhancing the efficiency and effectiveness of breeding programs (Xu et al., 2019). Livestock traits, particularly those related to productivity, health, and reproduction, are often complex and influenced by multiple genes with varying effects. These traits are typically categorized into two groups: qualitative traits, controlled by a few genes with large effects, and quantitative traits, influenced by many genes each contributing a small effect. For quantitative traits, which include most economically important characteristics such as milk yield, growth rate, and disease resistance, the genetic basis is polygenic. Advances in genomics have enabled the identification of SNPs associated with these traits, providing a deeper understanding of the genetic architecture underlying livestock performance. This knowledge is crucial for implementing GS, as it allows breeders to estimate the cumulative effects of numerous small-effect genes, thereby improving the accuracy of selection. The ability to assess the genetic potential of animals at an early age, without waiting for phenotypic data, represents a significant advancement over traditional breeding methods (Singh et al., 2019). 2.2 Theoretical framework of genomic selection The theoretical foundation of GS is based on the principle that all loci across the genome contribute to the genetic variance of a trait. In contrast to marker-assisted selection (MAS), which focuses on a few selected markers, GS considers the entire genome, including regions with small effects that might be overlooked in traditional approaches. The effectiveness of GS lies in its ability to use dense genetic marker data to predict breeding values with high accuracy. This is achieved through the use of statistical models that estimate the effects of all SNPs simultaneously, allowing for the capture of additive genetic variation. The integration of genomic information into breeding programs has led to significant improvements in genetic gain, as it reduces the generation interval and increases the reliability of breeding value predictions. The theoretical framework of GS also accounts for non-additive genetic effects, such as dominance and epistasis, which can further refine the accuracy of predictions. As genomic technologies continue to advance, the framework of GS is expected to evolve, incorporating more sophisticated models and larger datasets to enhance selection accuracy (Meuwissen et al., 2016). 2.3 Genomic prediction models Genomic prediction models are central to the implementation of GS. These models use genome-wide marker data to estimate the breeding values of animals, allowing breeders to make selection decisions based on genetic potential rather than phenotypic performance alone. Several models are used in GS, each with its strengths and weaknesses. The most common models include Best Linear Unbiased Prediction (BLUP), Bayesian methods (such as BayesA and BayesB), and machine learning approaches. BLUP models are widely used due to their simplicity and effectiveness in predicting additive genetic values (Gutierrez-Reinoso et al., 2021).
RkJQdWJsaXNoZXIy MjQ4ODYzNA==