A mixed model for analyses of data on multiple genetic markers

Area of Science:

• Quantitative genetics
• Animal breeding
• Genomic selection

Background:

• Accurate estimation of breeding values is crucial for genetic improvement.
• Genetic markers linked to important traits can enhance prediction accuracy.
• Existing methods may not fully account for complex pedigree structures or linkage disequilibrium.

Purpose of the Study:

• To develop a mixed model for analyzing genetic marker data to improve breeding value estimation.
• To provide a flexible model applicable to arbitrary pedigree structures in outbreeding species.
• To simplify the model for crosses between inbred lines, clarifying its relation to multiple regression.

Main Methods:

• A mixed model incorporating a relationship matrix for marked chromosome segments or QTL alleles.
• Utilizing a reduced animal model approach to minimize estimated effects.
• Implementing a grouping strategy to handle crossbreeding and linkage disequilibrium.

Main Results:

• The proposed mixed model accurately estimates breeding values using genetic marker data.
• The model effectively handles arbitrary pedigrees and accounts for crossbreeding and linkage disequilibrium.
• A simplified version of the model aligns with traditional multiple regression approaches for inbred line crosses.

Conclusions:

• The developed mixed model offers a powerful tool for enhancing the accuracy of breeding value estimation in animal populations.
• The model's flexibility makes it suitable for diverse breeding schemes, including outbreeding and crossbreeding.
• This approach integrates genomic information effectively into quantitative genetic analyses.