Molecular Plant Breeding 2025, Vol.16, No.5, 268-277 http://genbreedpublisher.com/index.php/mpb 270 3.2 Haplotype detection tools: sequencing platforms, variant calling, and haplotype phasing The detection of haplotypes mainly relies on high-throughput sequencing platforms, such as Illumina short-read sequencing, Oxford Nanopore and PacBio long-read sequencing. By combining variant calling and haplotype phasing algorithms, haplotype information can be obtained. Commonly used phase determination tools include EAGLE2, BEAGLE, SHAPEIT2, etc. Different tools perform differently under different data types and group structures. There are mainly two methods for dividing haplotype blocks: the method based on linkage disbalance (LD) and the sliding window method. The LD method is usually more accurate. Analyzing with multiple tools together can make the phase determination results more reliable. In addition, new methods such as HaploBlocker can adapt to data with different marker densities and genetic diversity by focusing on the linkage structure within the population (Otte and Schlotterer, 2020; Weber et al., 2023). 3.3 Characterizing haplotype diversity across global rice germplasm collections Large-scale sequencing projects such as 3K-RG and MiniCore have revealed the wide distribution of haplotypes in the rice genome and their specificity among different subspecies and regions. Most haplotypes are from specific subspecies or specific populations. In the modern breeding process, the frequency of some major haplotypes has undergone significant changes. There are also many unique haplotypes between wild rice and cultivated rice, some of which have been selected during domestication and adaptation. In-depth research on haplotype diversity provides a solid foundation for the exploration of superior alleles and molecular design breeding (Shang et al., 2022; Aung et al., 2024; Huang et al., 2024). 4 Haplotype-Trait Associations 4.1 Statistical models linking haplotypes to yield phenotypes Methods such as genome-wide association study (GWAS) and mixed linear models (MLM) can control population structure and kinship, reducing false positives. GWAS, combined with the best linear unbiased prediction (BLUP) values and multi-year phenotypic data, can effectively identify haplotype blocks and candidate genes related to yield. Some researchers used 2.8 million SNPS and BLUP values to detect 816 SNP signals significantly related to 13 agronomic traits in 259 rice materials, and identified candidate genes through haplotype block construction (Wang et al., 2020; Wang et al., 2021; Wang et al., 2023). Anandan et al. (2022) and Al-Daej et al. (2023) found that mixed linear models and unified mixed models were also used for association analysis, reducing the interference caused by group structure and kinship. 4.2 Multi-environment validation of haplotype effects Verifying the results in different environments is an important step to ensure the reliability of haplotype-trait associations. Studies have shown that some haplotypes or QTLs can significantly affect yield traits in different ecological environments, different years and different genetic backgrounds. For instance, GWAS was conducted on traits such as flowering period under field and greenhouse conditions in the United States, Bangladesh and the United Kingdom, and it was found that ten genomic regions were associated with candidate genes in one or more environments. A single SNP can explain 5% to 50% of phenotypic variations. Some QTLs have stable effects in different geographical environments, indicating that these haplotypes have strong environmental adaptability and breeding potential (Bharamappanavara et al., 2023). 4.3 Functional haplotypes vs. neutral haplotypes: prioritization for breeding Functional haplotypes refer to haplotypes that have a significant impact on the target trait. Such haplotypes are preferred in molecular design breeding. Studies have found that in modern rice varieties, the frequencies of favorable haplotypes of most known yield-related genes are relatively low, indicating that by mining and aggregating these favorable haplotypes, it is expected to significantly increase the yield (Zhang et al., 2021; Wang et al., 2023). The prioritization of functional haplotypes usually takes into account their interpretability for phenotypes, stability across environments, and association with known functional genes. For instance, some haplotypes have been precisely localized in near-isogenic lines, showing significant structural and expression differences, which directly affect the expression of yield QTLs. On the contrary, neutral haplotypes have no significant impact on traits, so they are not given priority in breeding. The identification and sequencing of functional haplotypes provide a theoretical and practical basis for high-yield molecular breeding of rice.
RkJQdWJsaXNoZXIy MjQ4ODYzNA==