Authors
Xiao-Lin Wu, Guilherme J M Rosa, Allicia Horn, Ezequiel Nicolazzi, John B Cole
Published in
JDS communications. Volume 7. Issue 6. Pages 863-868. Epub Oct 05, 2026.
Abstract
Binning continuous phenotypes into discrete categories is a longstanding practice in dairy cattle evaluation and management, yet its effects on genetic evaluation and selection remain insufficiently characterized. We develop a theoretical framework to investigate how binning affects heritability, EBV, and selection response. By expressing the discretized phenotype as a linear projection of the underlying continuous phenotype, we show that binning is governed by 2 quantities: an attenuation factor that scales the genetic signal and a variance rescaling term. These 2 components jointly determine the reduction in heritability, the degree of shrinkage in mixed-model equations, and EBV reliability. Coarse discretization, such as binary traits, results in substantial losses of genetic information and selection response. In contrast, moderate binning (e.g., 7-10 categories) retains most of the underlying signal, with only minor reductions in accuracy. Further increases in the number of categories yield diminishing returns, as genetic parameters and selection efficiency converge rapidly to those of the continuous trait. Selection response is relatively more robust to discretization than variance-based measures, indicating that individual rankings are largely preserved with a sufficient number of categories. These findings provide practical guidance for the design and interpretation of discretized traits in dairy genetic evaluations, highlighting the trade-off between recording simplicity and information retention.
PMID:
43016421
Bibliographic data and abstract were imported from PubMed on 12 Nov 2026.
Read full publication at:
Please sign in
to see all details.
Advertisement
Stats
- Recommendations n/a n/a positive of 0 vote(s)
- Views 3
- Comments 0