Quantile regression in genomic selection for oligogenic traits in autogamous plants: A simulation study

Area of Science:

• Plant breeding
• Quantitative genetics
• Statistical genomics

Background:

• Genomic selection (GS) is crucial for breeding autogamous plants with oligogenic traits.
• Traditional methods like RR-BLUP and BLASSO have limitations in accelerating allele fixation.
• Regularized Quantile Regression (RQR) offers a novel approach for genomic prediction.

Purpose of the Study:

• To evaluate the efficiency of RQR-based genomic selection for oligogenic traits in simulated autogamous plant populations.
• To compare RQR with traditional genomic selection methods (RR-BLUP, BLASSO).
• To identify statistical methodologies for the fastest fixation of favorable alleles.

Main Methods:

• Simulated F2 populations with varying heritability (0.10-0.40) and four controlling genes.
• Generations advanced to F6 under selection intensities of 10% and 20%.
• Genomic breeding values computed using RQR (quantiles 0.10, 0.50, 0.90) and compared with RR-BLUP and BLASSO.

Main Results:

• RQR generally performed equal to or better than traditional GWS methods.
• RQR achieved favorable allele fixation in most evaluated scenarios.
• RQR (0.50 quantile) rapidly fixed alleles by the fourth generation at 0.40 heritability and 10% selection intensity.

Conclusions:

• RQR is a promising tool for genomic selection in plant breeding of autogamous populations.
• Application of RQR can significantly reduce breeding time and costs by minimizing selfing generations.
• RQR facilitates faster fixation of favorable alleles for oligogenic traits.