Pre-selection of markers for genomic selection
Author Information
Author(s): Torben Schulz-Streeck, Joseph O. Ogutu, Hans-Peter Piepho
Primary Institution: University of Hohenheim
Hypothesis
Can pre-selecting markers improve the predictive accuracy of genomic breeding values?
Conclusion
Pre-selection of markers was beneficial and increased predictive accuracy from 0.607 to 0.625, with further improvements to 0.648 when partitioning markers into two groups with heterogeneous variances.
Supporting Evidence
- Pre-selecting markers increased the correlation between GEBVs and true breeding values from 0.607 to 0.625.
- Using an extended model with heterogeneous variances improved predictive accuracy to 0.648.
- Ridge regression and spatial models provided similar fits in the analysis.
Takeaway
Choosing the right markers before predicting breeding values helps make better predictions, like picking the best players for a team to win a game.
Methodology
The study used a simulated dataset and various methods including ridge regression and spatial models to evaluate the performance of pre-selected markers through cross-validation.
Limitations
The study may not apply to all scenarios, as pre-selecting markers can sometimes reduce accuracy.
Participant Demographics
3226 individuals across five generations, with 2326 phenotyped and genotyped individuals.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website