Abstract
BACKGROUND: Genome-wide association study (GWAS) is a strategy for genetic dissection of quantitative traits with high mapping resolution. Haplotypes based on multiple SNPs provide an effective alternative for exploring loci associated with complex quantitative traits. In this study, haplotype-based GWAS (Hap-GWAS) was conducted for seven fruit-related traits in the 287 tomato accessions, consisting of 237 S. lycopersicum, 30 S. lycopersicum var. cerasiforme, and 20 S. pimpinellifolium. RESULTS: Phenotypic variations of the fruit-related traits (weight, shape, locule number, pericarp thickness, number of flowers, number of fruits, and Brix) were assessed in three-years of field trials. Over 28.7 million SNPs were identified from whole genome-resequencing and a core set of 137,706 SNPs was used to construct 3,884 haplotype blocks across 12 chromosomes. The 11,970 haplotypes in these blocks found high levels of genetic diversity and differentiation with and between species in the GWAS panel. A total of 32 haplotypes, corresponding to 30 QTL, showed significant associations with fruit-related traits at a false discovery rate (FDR) adjusted P < 0.05 in at least two of the four phenotypic datasets (each of three years and combined), explaining up to 51.67% of the total phenotypic variance. Of these, 22 haplotypes were associated with 1-11 putative novel QTL across the seven traits and three haplotypes exhibited significant associations with multiple traits, suggesting pleiotropic effects of QTL. Furthermore, Hap-GWAS identified additional QTL that were not detected by SNP-based analysis, indicating the complementary power of this approach to capture multi-allelic and LD-based genetic effects underlying complex traits. CONCLUSIONS: These results support the effectiveness of Hap-GWAS in dissecting quantitative traits and facilitate development of an advanced breeding strategy for improving fruit-related traits in tomato.