Comparison of Genomic Selection Methods for Complex Traits
Author Information
Author(s): Pszczola Marcin, Strabel Tomasz, Wolc Anna, Mucha Sebastian, Szydlowski Maciej
Primary Institution: Poznan University of Life Sciences
Hypothesis
The study aims to compare different approaches used to calculate direct genomic values for quantitative and binary traits.
Conclusion
Bayesian model averaging methods predicted breeding values slightly better than GBLUP, and methods using genomic information outperformed traditional pedigree-based analyses.
Supporting Evidence
- Accuracy for quantitative traits ranged from 0.26 to 0.89 for males and 0.31 to 0.89 for females.
- Bayesian methods generally performed better than traditional methods.
- Bivariate analyses were slightly advantageous over single trait analyses.
Takeaway
Scientists tested different ways to predict animal traits using genetic data, and found that some methods worked better than others.
Methodology
Participants applied 26 approaches for quantitative traits and 15 for binary traits to estimate direct genomic values from simulated data.
Limitations
None of the methods estimated the non-additivity of QTL affecting the quantitative trait, which may limit accuracy.
Participant Demographics
The study involved 11 groups of researchers who submitted their estimates of genomic breeding values.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website