A chromosome-level genome assembly of an elite indica breeding line resolves the genomic architecture of drought-yield QTLs qDTY2.2 and qDTY4.1 in rice.

Drought-yield quantitative trait loci (QTLs) qDTY2.2 and qDTY4.1 improve grain yield under reproductive-stage drought in rice (Oryza sativa L.), yet their genomic architecture remains poorly resolved in elite indica breeding backgrounds. We generated a chromosome-level assembly of IR87707-446-B-B-B (IR64-qDTY2.2 + qDTY4.1), which carries both QTLs in the IR64 background, and delineated donor-derived introgressions using informative parental single-nucleotide polymorphisms (SNPs). The 395.5-Mb assembly comprised 12 chromosomes, showed 99.2% BUSCO completeness, and contained 40,189 predicted protein-coding genes. Informative-SNP analysis, supported by read-depth assessment of low-information regions, resolved qDTY2.2 and qDTY4.1 as continuous donor-derived introgressions containing 194 and 97 genes, respectively. Comparative genome analysis showed that qDTY2.2 was largely macrostructurally collinear with IRGSP-1.0, whereas qDTY4.1 showed pronounced structural divergence and contained a 15.6-kb inversion relative to IR64sj. Comparative gene-content analysis further showed that 37.1% and 55.7% of genes in qDTY2.2 and qDTY4.1, respectively, lacked collinear orthologs in IRGSP-1.0. Multi-accession allele analysis revealed contrasting candidate-gene profiles between the two loci. Six of eight qDTY2.2 candidates had coding sequences identical to IR64sj, whereas all four qDTY4.1 candidates showed coding differences from IR64sj. Coordinate-based SNP markers were also developed for donor-haplotype tracking across both QTL intervals. These results provide a genome-resolved framework for characterizing drought-yield QTLs and developing breeding-background-specific markers for recombination-aware rice breeding.