Legume Genomics and Genetics 2026, Vol.17, No.1, 49-67 http://cropscipublisher.com/index.php/lgg 61 the initial population while reducing the number of genotypes or markers offer additional cost savings without compromising accuracy, particularly when model-based strategies are used to choose training individuals and markers (Ćeran et al., 2024). With these methodological and technological advances, SNP-enabled GS is poised to become a routine component of soybean breeding programs, supporting rapid recycling of parents, optimal cross prediction, and multi-trait selection for yield, quality, and adaptation (Jiahao et al., 2025). 7.3 Multi-omics technologies and soybean genetic improvement The expanding use of SNP markers is increasingly integrated with other omics layers—transcriptomics, metabolomics, and phenomics—to dissect complex traits and drive more precise soybean improvement. High-density SNParrays and resequencing provide the foundational genomic variation, which can be linked to expression (eQTL) data, metabolite profiles, and detailed phenotypes to identify causal genes and pathways underlying yield, stress tolerance, and seed quality (Miller et al., 2023; Gai et al., 2025). For example, combining GWAS with transcriptomic data has helped prioritize candidate genes within QTL regions for traits such as flowering and maturity, while metabolite-associated SNPs refine the genetic control of nutritional components like sucrose, isoflavones, tocopherols, and amino acids (Jiahao et al., 2025). Multi-omics data also reveal pleiotropic effects and gene networks, providing systems-level targets for breeding and genome editing rather than focusing solely on single loci. In parallel, advances in high-throughput genotyping (e.g., GBTS panels, GBS) and phenotyping, together with machine-learning and artificial-intelligence approaches, are reshaping how breeders exploit SNP-based diversity in a multi-omics context. Integrative frameworks now couple SNP-based genomic prediction with environmental, physiological, and management data to model genotype-by-environment interactions and support climate-resilient cultivar development (Miller et al., 2023). Multi-omics-informed GS models, which incorporate SNPs alongside expression or metabolite markers, are being explored to improve prediction for complex traits and to design ideotypes optimized for both productivity and sustainability. At the same time, large resequencing datasets and SNP resources spanning thousands of accessions facilitate the construction of mutant gene libraries and enable rapid identification of natural alleles suitable for CRISPR/Cas-based editing or allele replacement. Together, these developments indicate that SNP markers will remain the genomic backbone of soybean improvement, increasingly embedded within holistic, multi-omics breeding strategies that integrate MAS, GS, and genome editing to accelerate genetic gain and broaden the adaptive potential of global soybean germplasm. 8 Summary and Outlook Over the past decade, SNP genotyping and resequencing have transformed understanding of global soybean genetic diversity and population structure. Large-scale efforts have characterized hundreds to thousands of accessions spanning wild relatives, landraces, and elite cultivars, revealing millions to tens of millions of SNPs and providing a high-resolution view of genome-wide variation. These studies consistently distinguish wild from cultivated groups, identify transitional or hybrid genotypes, and show that domestication and modern breeding have dramatically reduced diversity and reshaped linkage disequilibrium (LD) through selective sweeps. At regional scales, population-wide SNP analyses have clarified how breeding history and geography structured germplasm in Brazil, Southern Africa, Kazakhstan, and Korea, often revealing a small number of genetic clusters and strong within-population variation with relatively low differentiation among groups. Such insights have directly informed strategies to broaden the genetic base and guide parental selection for adaptation to tropical, temperate, or stress-prone environments Parallel advances in SNP array and GBS platforms (e.g., SoySNP50K, Axiom SoyaSNP, DArT-SNP, and custom low-to medium-density panels) have delivered robust, evenly distributed, and cost-effective marker sets that are now widely used for diversity analysis, fingerprinting, and association mapping. Consolidation of resequencing data for 1.5K and 3,661+ accessions into unified variant resources with 30–32 million SNPs has created versatile public datasets for post-genomic research, facilitating in-silico genotyping, high-resolution GWAS, and cross-collection comparisons. These genomic resources underpin discovery of loci for domestication traits, yield components, quality, and stress tolerance, and enable identification of region-specific or rare favorable alleles in
RkJQdWJsaXNoZXIy MjQ4ODYzNA==