Effects of Allele Coding Methods on Genomic Evaluation
Author Information
Author(s): Ismo Strandén, Ole F. Christensen
Primary Institution: MTT Agrifood Research Finland and Aarhus University
Hypothesis
Different allele coding methods will lead to the same inference in genomic evaluations when a fixed general mean is included in the model.
Conclusion
Different allele coding methods lead to the same inference in marker-based and equivalent models when a fixed general mean is included, but they affect the reliabilities of genomic breeding values.
Supporting Evidence
- Theoretical derivations showed that parameter estimates are the same irrespective of allele coding when a fixed general mean is included.
- Reliabilities of estimated genomic breeding values depend on the allele coding method used.
- The centered allele coding method showed better mixing properties in Markov chain Monte Carlo methods.
Takeaway
This study looked at how different ways of coding genetic information affect the accuracy of predictions in animal breeding. It found that while the predictions can be similar, the reliability of those predictions can change based on the coding method used.
Methodology
The study used theoretical derivations and practical examples to analyze the effects of different allele coding methods on genomic breeding values using restricted maximum likelihood and Bayesian methods.
Limitations
The results are based on the assumption that phenotypes and genotypes are available for all animals in the analysis, which may not always be the case.
Participant Demographics
The analyzed data included 4665 animals with phenotypes across four generations.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website