Computational Molecular Biology 2024, Vol.14, No.2, 54-63 http://bioscipublisher.com/index.php/cmb 57 facilitating the seamless integration of diverse data types. Bioinformatics pipelines are essential for processing and interpreting GBS datasets, enabling the identification of genetic markers and the development of GS models (Ganal et al., 2014; He et al., 2014). Additionally, databases that compile marker data from multiple genotyping experiments can streamline downstream data processing and enhance the utility of genotyping data for both scientific research and breeding applications (Ganal et al., 2014). Figure 1 The role of NGS-Based marker technology and high-throughput phenotyping in genomic selection (Adopted from Bhat et al., 2016) 4 Statistical Methods and Models 4.1 Best linear unbiased prediction (BLUP) Best Linear Unbiased Prediction (BLUP) is a widely used statistical method for estimating random effects in mixed models, particularly in the context of animal breeding. Originally developed for estimating breeding values, BLUP has been adapted for various applications, including plant breeding and variety testing. In plant breeding, BLUP has been employed to model and exploit genetic correlations among relatives using pedigree information, and to handle genotype-by-environment interactions through flexible variance-covariance structures. This method has demonstrated good predictive accuracy compared to other procedures, making it a valuable tool for genetic evaluation in both plants and animals (Piepho et al., 2008). In animal breeding, BLUP has been adapted to address specific challenges such as the inclusion of dam effects in models for polytocous species like swine and poultry. This adaptation involves hierarchical models that account for sires, dams within sires, individuals within full-sib families, and records within individuals. The development of alternative computing algorithms has facilitated the timely genetic evaluation of large populations, ensuring that BLUP remains a robust and efficient method for genetic prediction. Additionally, the integration of genomic information into BLUP models, such as the use of trait-specific marker-derived relationship matrices, has further enhanced the accuracy of genomic breeding value predictions (Bauer et al., 2006; Muir, 2007; Zhang et al., 2010). 4.2 Bayesian methods Bayesian methods have gained prominence in genomic prediction due to their flexibility and ability to incorporate prior information. These methods, such as BayesA, BayesB, and BayesC, allow for the estimation of marker
RkJQdWJsaXNoZXIy MjQ4ODYzNA==