Predicting genomic selection efficiency to optimize calibration set and to assess prediction accuracy in highly...

Area of Science:

• Plant breeding
• Quantitative genetics
• Genomics

Background:

• Genomic selection (GS) uses genotypic data to predict selection candidate performance.
• Prediction accuracy in GS is influenced by calibration set (CS) composition.
• Existing criteria for CS optimization are less effective for structured populations common in plant breeding.

Purpose of the Study:

• To develop and evaluate criteria for predicting genomic selection efficiency in structured populations.
• To optimize calibration set sampling and estimate prediction reliability for multiparental populations.
• To adapt existing theories for structured breeding materials.

Main Methods:

• Derived criteria from generalized coefficient of determination (CD) theory.
• Applied criteria to optimize CS sampling and assess prediction reliability.
• Evaluated methods on maize Nested Association Mapping (NAM) populations and diverse panels.

Main Results:

• The proposed CD criteria were efficient for sampling optimized CS in most tested scenarios.
• Criteria partially estimated prediction reliability between NAM families.
• The criteria could not fully differentiate reliability when using diverse panels as CS for NAM families.

Conclusions:

• The developed CD criteria are adaptable to various prediction scenarios, including inter- and intra-family predictions.
• These criteria improve prediction accuracies in structured populations.
• The criterion is valuable for defining optimal calibration sets and estimating prediction reliability.