The Camellia sinensis var. sinensis cv. Fuding Dabaicha genome unveils structural variation-driven metabolic innovation

Tea plants possess a highly heterozygous genome and produce diverse beneficial metabolites, yet the genomic basis of its metabolic diversity remains fully elusive. Here, we show that single-cell sequencing of 107 sperm cells, combined with PacBio HiFi and ONT ultra-long sequencing, enables an accurate haplotype-resolved genome assembly of Fuding Dabaicha (FDDB). Structural variations (SVs) between the two haplotypes comprise 23.8% of the genome and strongly influence crossover patterns. Using these phased genomes, we establish a half-sib-based QTL mapping platform and identify a Gypsy LTR insertion in the promoter of CsDFRb, which associated with the increasing of p-coumaroylquinate levels in young leaves. Moreover, mGWAS reveals 2649 additional loci associated with 2837 metabolites when using the FDDB genome rather than the 'Tieguanyin' reference genome for variant calling. Functional validation of CsC3H and CsST2Ac confirms their role in determining chlorogenic acid and sulfated metabolite levels in tea plant. In addition, we observe that allelic heterogeneity at CsST2Ac affects sulfated metabolite abundance. These phased genomes illuminate how SVs drive metabolic diversity and offer valuable resources for tea breeding.