068
Hierarchical Quantitative Genetic Model Using Genomic Information

Wednesday, August 20, 2014: 10:45 AM
Bayshore Grand Ballroom B-C (The Westin Bayshore)
Gregor Gorjanc , The Roslin Institute and R(D)SVS, University of Edinburgh, Midlothian, United Kingdom
John A. Woolliams , The Roslin Institute and R(D)SVS, University of Edinburgh, Midlothian, United Kingdom
John M. Hickey , The Roslin Institute and R(D)SVS, University of Edinburgh, Midlothian, United Kingdom
Abstract Text:

Genomic evaluations are commonly based on models with genomic relationships. However, estimated genomic relationship matrices are often positive semi-definite and ad-hoc corrections are applied to force positive definiteness without consideration about model properties. In this contribution a hierarchical quantitative genetic model is postulated that provides a positive definite genomic relationship matrix by taking into account the amount of genetic variance captured by estimated marker effects. Based on the hierarchical formulation the proposed model also provides a system of equations to estimate marker effects and breeding values jointly without setting up and inverting covariance matrices. Further extension with pedigree information is also possible.

Keywords:

genomics

hierarchical model

inverse