A graph model for genomic prediction in the context of a linear mixed model framework

Area of Science:

• Agricultural Science
• Genetics
• Bioinformatics

Background:

• Genomic selection (GS) is crucial for advancing plant and animal breeding.
• High prediction accuracy is essential for the successful application of GS.
• Linear mixed models (LMMs) are commonly used in genomic prediction.

Purpose of the Study:

• To enhance prediction accuracy in genomic selection.
• To investigate the utility of graph models within a linear mixed model framework.
• To evaluate the impact of incorporating line connections and genotype effects.

Main Methods:

• Exploration of graph models integrated into a linear mixed model framework.
• Comparison of prediction accuracy using genotype effects alone, graph structure alone, and combined effects.
• Validation across 14 diverse plant breeding datasets.

Main Results:

• Genomic prediction accuracy decreased when only graph structure (line connections) was used.
• Integrating both genotype effects and graph structure yielded a slight improvement over using genotype effects alone.
• These results were consistent across multiple datasets.

Conclusions:

• Graph models alone do not enhance genomic prediction accuracy in this context.
• Combining genotype effects with graph structure offers a marginal improvement in prediction accuracy.
• The findings provide insights for optimizing genomic selection strategies in plant breeding.