Genome-wide SNP discovery through genotyping by sequencing, population structure, and linkage disequilibrium in Brazilian peach breeding germplasm

Genotyping by sequencing (GBS) is a flexible and cost-effective strategy for genome-wide SNP discovery and high-throughput genotyping. Here, we employed the GBS approach to explore population structure, genetic variability, and patterns of linkage disequilibrium (LD) among 220 peach genotypes (Prunus persica), representative of the Brazilian breeding germplasm. This selected panel is mainly composed by locally adapted peaches developed in mild winter and high relative humidity conditions, and represents a worldwide reference of peach germplasm for low-chilling areas. A total of 93,353 SNP markers were discovered, and after filtering, 18,373 high-quality SNPs were used in analyses. Thirty-four percent of selected SNPs were located in genic regions and about 70% of these in the coding sequence. Principal coordinate analysis (PCoA) and Bayesian clustering (fastSTRUCTURE) successfully detected population genetic structure in our germplasm panel, supporting the identification of three main distinct subpopulations. The distribution of the genotypes within subpopulations reflected their fruit-related traits: melting and non-melting flesh cultivars. LD patterns revealed a medium LD level in peach, with the extent of LD highly dependent on the subpopulation and genome regions. The use of GBS-SNPs to perform genome-wide association studies (GWAS) was also exploited by using phenotypic information from five fruit quality traits. The observed significant SNP-trait associations were always in the regions previously predicted by linkage and association mapping analysis, providing valuable insights to incorporate the use of marker-assisted selection in the breeding program.