Abstract
BACKGROUND: Fruit quality traits such as fruit weight, firmness, total soluble solids content, and color strongly influence consumer acceptance, market value, postharvest shelf life, and processing efficiency in tomato. Therefore, these traits are central to breeding programs but remain challenging to dissect due to their polygenic and pleiotropic nature. RESULTS: We evaluated 167 accessions from the Varitome collection, encompassing Solanum pimpinellifolium, Solanum lycopersicum var. cerasiforme, and Solanum lycopersicum var. lycopersicum. The Varitome collection represents a rich source of genetic and phenotypic diversity, making it a powerful resource for mapping complex traits. Genome-wide association studies (GWAS) were conducted using the FarmCPU and Blink models, utilizing a dataset of 3,879,252 SNPs, 831,152 INDELs, and 11,447 structural variants (SVs). Six fruit-quality traits were phenotyped: fruit weight, firmness, total soluble solids, lightness (L*), chroma (C*), and hue (h°). Our multi-variant GWAS uncovered both known and novel determinants of fruit quality. Known loci such as PSY1 and fw11.3/CSR were validated, while robust new signals were detected on chromosomes 1, 4, 6, 8, and 12 for °Brix and on chromosome 7 for fruit weight. Several pleiotropic hotspots were identified, particularly on chromosomes 1, 5, 6, and 8 for fruit color, supported by convergent SNP, INDEL, and SV associations. Candidate genes included biosynthetic enzymes (PSY homologs, LIN5), sugar transporters (SWEETs, SUTs, and sugar facilitator proteins), transcriptional regulators (MADS-box, bHLH, TCP, NAC, and MYB families), and genes linked to plastid remodeling, light signaling, and oxidative turnover. Integration of INDELs and SVs across models improved mapping resolution and robustness, enabling the detection of loci that would remain hidden in SNP-only scans. CONCLUSIONS: This study demonstrates the multilayered genetic networks governing tomato fruit quality and expands the catalog of loci contributing to polygenic traits. By using SNPs, INDELs, and SVs with FarmCPU and Blink models, we provide validated and novel targets for marker-assisted breeding, genomic selection, and functional validation. These findings establish a framework for accelerating the development of tomato cultivars with enhanced fruit weight, sweetness, firmness, and color, thereby supporting both market competitiveness and nutritional quality.