|AKDEMIR, DENIZ - Cornell University
Submitted to: Genetics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 1/2/2015
Publication Date: 3/1/2015
Publication URL: http://DOI: 10.1534/genetics.114.173658
Citation: Akdemir, D., Jannink, J. 2015. Locally epistatic genomic relationship matrices for genomic association. Genetics. 199:857-871.
Interpretive Summary: In plant and animal breeding studies a distinction is made between the genetic value of an individual and its breeding value. The latter contains only additive effects that are transmitted from parents to offspring, while the former contains interaction effects that will not be transmitted if recombination occurs. We argue that breeders can take advantage of some interaction effects that occur in regions of low recombination. We introduce models to estimate genetic effects caused by such local interactions by using genetic map information and combining the local additive and interaction effects. The models estimate the relationship among individuals locally, in defined chromosomal segments, rather than, as is currently common, genome-wide. Because the models evaluate relationships in many segments, they have high dimension and need to be post-processed to generate a simple and interpretable final model. Our models produce good predictive performance along with good explanatory information.
Technical Abstract: In plant and animal breeding studies a distinction is made between the genetic value (additive + epistatic genetic effects) and the breeding value (additive genetic effects) of an individual since it is expected that some of the epistatic genetic effects will be lost due to recombination. In this paper, we argue that the breeder can take advantage of some of the epistatic marker effects in regions of low recombination. The models introduced here aim to estimate local epistatic line heritability by using the genetic map information and combine the local additive and epistatic effects. To this end, we have used semi-parametric mixed models with multiple local genomic relationship matrices with hierarchical designs and lasso post-processing for sparsity in the final model. Our models produce good predictive performance along with good explanatory information.