Evaluation of genome-wide selection efficiency in maize nested association mapping populations

Area of Science:

• Plant Breeding
• Genomics
• Quantitative Genetics

Background:

• Marker-assisted selection (MAS) uses a subset of genetic markers, potentially missing quantitative trait loci (QTL) with small effects.
• Genome-wide selection (GWS) utilizes all available genetic markers for improved prediction of breeding values (BVs).

Purpose of the Study:

• To evaluate the prediction accuracy of GWS for three maize (Zea mays) flowering traits.
• To compare GWS performance against conventional marker-assisted selection (MAS).

Main Methods:

• Cross-validation experiments were conducted on a large dataset from 25 maize nested association mapping populations.
• Genome-wide selection was implemented using ridge regression-best linear unbiased prediction (RR-BLUP).
• Comparison was made with composite interval mapping (CIM) and other GWS methods (BayesA, BayesB).

Main Results:

• RR-BLUP significantly outperformed MAS utilizing composite interval mapping (CIM) for predicting maize flowering traits.
• RR-BLUP demonstrated prediction accuracies comparable to or greater than BayesA and BayesB.
• Prediction accuracy with RR-BLUP increased with training sample proportion, marker density, and heritability, eventually reaching a plateau.

Conclusions:

• GWS, particularly RR-BLUP, offers superior prediction accuracy for complex traits in maize compared to traditional MAS.
• RR-BLUP is a preferred method for estimating marker effects in GWS, showing robustness across various factors.
• Gains in accuracy with RR-BLUP over CIM were generally larger when training sample proportion, marker density, and heritability were lower.