EPDs posted from the 11/18 run included evaluation changes to Calving Ease (CE) & Maternal Calving Ease (MCE)
What does this mean?
IGS made three improvements that make Calving Ease (CE) & Maternal Calving Ease (MCE) predictions cleaner, more accurate, and more consistent. With change comes some re-ranking and additional spread from highest to lowest values. The information IGS released is at the bottom of this post, and here’s the Data Genie interpretation & breakdown.
Breed and Decade Interaction Effect Updates
– More stable, consistent EPDs
– Less noise when a breed or time period has minimal data
– Will allow for cleaner comparisons across years and across different breeds in the IGS system
External EPD Incorporation
– More predictable calving ease trends in a multi-breed evaluation with considerations of internal and external information
Marker Effects
– CEM is now a genomic EPD which means we’ll see higher accuracy for this trait
– Improved predictions for MCE
– Updated handling of relationships between related traits
Summed Up
Model improvements always cause some level of movement. We trust that IGS performed their validation process effectively and that this update will better characterize the genetic potential of animals for calving ease (direct and maternal). So far, with this update, we are noticing a shift in the CE mean coupled with an increased range of EPDs for many animals CE and to a lesser degree MCE EPD values moved lower, even though their relative % ranking within the breed may be similar to before. If anything appears inconsistent or doesn’t line up with what you expect in your herd, please flag it and we can help review, and work with the association on explanations as needed.
I know it’s tough while we’re living through one, but evaluation updates are a necessary and healthy part of genetic progress. As new data, methodology, and computing power becomes available, the IGS system will continue to evolve and that’s a good thing. There’s rarely an ideal time to release updates, but we have to work with what we’re given. The focus now should be reviewing your herd specific data and using percentile ranks to help interpret any changes, alerting if something is out of sync. If you have questions on specific animals or need a report sent to help evaluate your animals, and are a client, please reach out to us because that’s exactly what Data Genie is here for.
Adapted from: ASA (https://herdbook.org/simmapp/template/epdAnalytics%2CPurebred.vm and https://herdbook.org/simmapp/template/epdAnalytics%2CHybrid.vm)
Updates to CED and CEM Genetic Evaluation – from IGS
International Genetics Solutions (IGS) has updated the calving ease direct (CED) and calving ease maternal (CEM) genetic evaluation. These updates have increased the efficiency and prediction accuracy of the evaluation. Three updates were made as outlined below.
Breed and Decade Interaction Effect. Previously, a breed and decade interaction was fit as a random effect. Some breed x decade groups had sparse representation causing decreased evaluation efficiency. Now, breed and decade effects are decoupled, or fit as two separate effects.
External EPD Incorporation. Previously, external EPD were included directly in the coefficient matrix. Now, external EPD are included after the internal EPD analysis. This will allow the impact of external EPD information to be more clearly compared.
Marker Effects. Previously, marker effects for maternal effects were not included. Now, methodology has been developed that has the ability to fit marker effects for direct and maternal effects simultaneously. CEM EPD are now genomically enhanced. In addition, updates to the covariances between marker effects for different traits included in the evaluation were also updated.
These updates have undergone an extensive validation process and prediction accuracy of the CED and CEM genetic evaluation has improved. Members may see slight re-ranking as a result of these changes. Members with any questions regarding these updates or specific animals’ EPD can reach out to the breed Association staff or board members for more information.
datageniellc.com 