Dissecting seed oil content QTL and integrating their genetic effects for genomic prediction in Brassica napus

Abstract

Seed oil content (SOC) is a key determinant of oil yield in rapeseed, but translating high-resolution QTL and functional gene information into effective breeding selection remains challenging. Here, we integrated high-quality genome assembly, QTL fine mapping, gene function validation, and QTL-informed genomic prediction to improve the SOC in rapeseed. The improved, chromosome-scale genome of the semi-winter cultivar NY7 served as a reliable reference for fine mapping via its bidirectional introgression populations. Seven major QTLs were rapidly fine-mapped into 117 kb ~ 358 kb intervals; each increased the SOC by 2% ~ 6%. Integrated transcriptomic and haplotype analyses revealed eight candidate genes, highlighting BnaDIR1.C2 as the hub gene underlying qOC.C2–1. Functional validation confirmed the positive regulation of BnaDIR1.C2 in SOC. Moreover, genomic prediction models incorporating QTL-weighted markers substantially improved the prediction performance by an average of 20.34% across different populations. This study bridges the gap between high-resolution genetic dissection and predictive breeding, providing a practical framework to accelerate oil yield improvement in rapeseed.