Single- and multiple-trait quantitative trait locus analysis for seed oil and protein contents of soybean populations with advanced breeding line background

Soybean seed oil and protein contents are negatively correlated, posing a challenge in breeding efforts to enhance both traits. Previous studies have identified hundreds of oil and protein QTLs mainly via the single-trait QTL analyses. Multiple-trait QTL methods for correlated traits were shown to improve detection power and mapping precision compared to single-trait methods but have not been applied to seed oil and protein contents. Our study conducted both single- and multiple-trait multiple interval mapping (ST-MIM and MT-MIM, respectively) for oil and protein contents using three recombinant inbred line populations derived from crosses between advanced breeding lines and tested in four environments. We detected seven ST-MIM QTLs on chromosomes 1, 8, 6, 15, 19, and two on 20, five of which were confirmed by MT-MIM. Using our original data as well as simulated data, our findings show the multiple-trait method did not outperform the single-trait approach for our traits of interest with high heritability (over 0.84). All loci exerted opposite effects on oil and protein contents, but the protein-to-oil additive effect ratio varied (-0.6 to -48.8). We calculated the allelic effects on estimated processed values (EPV) using the National Oilseed Processors Association (NOPA) and High Yield + Quality (HY+Q) methods. Oil-increasing alleles of QTLs on chromosomes 6, 15, 19, and 20 increased both EPV NOPA and EPV HY+Q, while oil-increasing alleles of QTLs on chromosomes 1 and 8 increased EPV NOPA and decreased EPV HY+Q, which penalizes low protein meal. With the populations' elite pedigree, selected lines can be used to determine the allelic effects on yield and directly integrated into breeding programs.