Molecular Plant Breeding 2024, Vol.15, No.3, 132-143 http://genbreedpublisher.com/index.php/mpb 134 3.1.2 Genomic selection Genomic selection (GS) represents a significant technological advancement in MAS, utilizing genome-wide markers to predict complex phenotypes. This approach has the potential to accelerate breeding cycles, increase selection intensity, and improve the accuracy of breeding values. GS has been particularly impactful in tree breeding, where long breeding cycles and complex genetic architectures pose significant challenges. Studies have shown that GS can outperform traditional MAS, especially when large populations are genotyped, and advanced computational models are employed (Grattapaglia et al., 2018; Sandhu et al., 2022; Degen and Müller, 2023). The study by Degen and Müller (2023) compared the effectiveness of advanced marker-assisted selection (MAS) and the widely used genomic selection (GS) in tree breeding programs. The research utilized a new software, "SNPscan breeder", to simulate simple tree breeding procedures and compare the impact of different selection criteria on genetic gain and inbreeding. The results indicated that GS outperformed other methods in nearly all simulated scenarios, especially when using the gBLUP method. MAS based on GWAS results only surpassed GS when the population used to estimate allele effects was very large (around 10 000 individuals) and the individuals had no kinship. Furthermore, GS increased inbreeding more than progeny testing and GWAS-based selection, which led to a stronger reduction in genetic diversity. The article discussed the practical implications of these findings for tree breeding programs and summarized the potential of GS in forest breeding and improvement, noting that MAS could become more relevant in the future as sequencing costs decrease (Figure 1). Figure 1 Genetic gains achieved by different selection criteria over 5 generations in 3 simulated breeding populations (Adopted from Degen and Müller, 2023) Image caption: a-c) Average cumulative genetic gains (n = 10) over 5 breeding cycles are shown for selection based on gBLUP, i.e. GS (blue), GWAS, i.e. MAS (green and yellow), phenotypes (orange), and progeny tests (red); d-f) Average genetic gains (n = 10) per breeding cycle (generation) are shown for the different selection criteria. Error bars represent the standard deviation of the mean; the 3 simulated breeding populations exhibit different genetic architectures, with 200 a, b, d, e) vs 20 c, f) causal variants, and different levels of kinship, with unrelated individuals a, d) vs average kinship values of approximately 0.04 b, c, e, f) (Adopted from Degen and Müller, 2023) Figure 1 compares genetic gains under different selection criteria over five breeding cycles. Different colors distinguish genomic selection (GS, blue), marker-assisted selection based on GWAS (MAS, green and yellow), phenotypic selection (orange), and progeny testing (red). The results indicate that selection based on progeny testing achieved the highest genetic gains across all test scenarios, with over 20% gain in the first breeding cycle and cumulative gains ranging from 40% to 70% over five breeding cycles. In contrast, MAS based on different GWAS analyses performed the worst, resulting in the lowest genetic gains.
RkJQdWJsaXNoZXIy MjQ4ODYzMg==