Computational Molecular Biology 2024, Vol.14, No.2, 54-63 http://bioscipublisher.com/index.php/cmb 60 considering all markers linked to genes affecting the trait, thus improving the accuracy of selection. The development of high-throughput genotyping technologies and the discovery of numerous single nucleotide polymorphisms (SNPs) have facilitated the widespread adoption of GS in livestock breeding programs (Meuwissen et al., 2016). This approach has been particularly impactful in dairy and beef cattle, pigs, and poultry, where it has significantly enhanced genetic gain by reducing the generation interval and increasing selection accuracy (Ibtisham et al., 2017). 6.2 Selection for reproductive traits Reproductive traits are crucial for the efficiency and profitability of livestock production. Genomic selection has shown promise in improving these traits by enabling early and accurate prediction of breeding values. By using genetic markers spread across the entire genome, GS can capture the effects of multiple quantitative trait loci (QTL) associated with reproductive performance (Ibtisham et al., 2017). This allows for the selection of animals with superior reproductive traits at a younger age, thereby accelerating genetic progress. The integration of GS with other breeding tools and platforms can further enhance the selection process, making it more efficient and cost-effective (Xu et al., 2019). 6.3 Health and productivity enhancements Improving the health and productivity of livestock is a primary goal of breeding programs. Genomic selection has been instrumental in achieving these objectives by providing a more precise estimation of genetic merit for health and productivity traits. Studies have shown that GS can significantly enhance the genetic gain for traits such as disease resistance, growth rate, and milk production (Ibtisham et al., 2017). The use of whole-genome regression models, which regress phenotypes on thousands of markers simultaneously, has been particularly effective in predicting complex traits (Campos et al., 2013). Additionally, the detection of selection signatures in livestock genomes has provided insights into the domestication and evolutionary processes, helping identify candidate genes associated with economically important traits (Saravanan et al., 2020). This knowledge can be leveraged to develop breeding strategies that improve the overall health and productivity of livestock populations. 7 Challenges and Limitations 7.1 Genotype-environment interactions Genotype-environment (G×E) interactions present a significant challenge in genomic selection for both plant and animal breeding. These interactions can complicate the prediction of phenotypic traits because the performance of genotypes can vary across different environments. Several studies have highlighted the importance of incorporating G×E interactions into genomic prediction models to improve accuracy. For instance, models that account for G×E interactions have shown superior predictive ability compared to single-environment models (Malosetti et al, 2016; Cuevas et al., 2016; Oakey et al., 2016). Additionally, the use of environmental covariables has been found to be beneficial in predicting phenotypes in untested environments, further emphasizing the need to consider G×E interactions in genomic selection (Malosetti et al, 2016). However, the complexity of these models and the computational resources required to implement them can be substantial, posing a significant limitation (Granato et al., 2018; Jighly et al., 2021). 7.2 Computational and resource constraints The implementation of genomic selection models, especially those incorporating G×E interactions, often requires significant computational resources. The scale of multi-environment trials is increasing, which in turn increases the computational challenges associated with genomic selection (Granato et al., 2018). For example, Bayesian models that account for G×E interactions can be computationally intensive, although recent advancements have led to more efficient algorithms and software packages that reduce computational time (Cuevas et al., 2016; Granato et al., 2018). Despite these advancements, the need for high computational power and extensive data storage remains a barrier, particularly for smaller breeding programs with limited resources (Lado et al., 2016; Jighly et al., 2021). Additionally, the integration of high-throughput sequencing data and the need to process large datasets further exacerbate these computational challenges (He et al., 2023).
RkJQdWJsaXNoZXIy MjQ4ODYzNA==