July 9, 2020
November 30, 2017
1. The Genomic Component
Today’s evaluation utilizes Molecular Breeding Values (MBVs) to predict the genomic contribution. This post evaluation, blending exercise, correlates the genetic relationship between the MBVs and the traits of interest. The updated genetic evaluation can more precisely evaluate the genomic differences between animals. This single step analysis evaluates Single Nucleotide Polymorphisms (SNPs) markers on cattle DNA using an analytical approach called the Marker Effects Model. The set of markers used have been selected to be the most informative for the traits in the genetic evaluation. Like phenotypes, the marker effects of a genotype directly impact an animal’s Expected Progeny Difference (EPD).
2. Model Changes
AHA has long conducted a full multi-trait genetic evaluation, meaning all traits are correlated to one another either through a favorable or unfavorable relationship. With the implementation of the new evaluation, traits have been grouped together into meaningful subsets that are optimal for inclusion in the Marker Effects Model analysis. The AHA implemented the de-coupling of models to more efficiently perform more frequent evaluations. Plus, moving away from a full-multi-trait evaluation will better allow for estimating traits of interest rather than building these traits through correlation as was done in the older analysis. The models are listed below.
2a. Birth Weight, Weaning Weight, Yearling Weight and Maternal Milk
The main growth model will stay the same with little impact to traits in the model. However, because of the new analysis use of the single step Marker Effects Model some animals’ EPDs will change. These changes will primarily be the result of increased accuracy using the new methods and reduced bias.
2b. Scrotal Circumference and Weaning Weight
Scrotal circumference is correlated with weaning weight, but only scrotal circumference EPDs will be reported out of this model. There is a decrease of correlated data impacting scrotal EPDs but this impact will be very small. Again, because of the change in the way the updated genetic evaluation is handling the genomic component there are changes in some animals’ scrotal circumference EPDs.
2c. Birth Weight, Ultrasound Back Fat, Ultrasound Intramuscular Fat, Carcass Weight, Carcass Back Fat, and Carcass Marbling Score
Birth weight is used as the correlated trait with the carcass traits to account for selection that may occur. Weaning and yearling weight is no longer included as correlated traits with carcass fat and carcass marbling. Instead scan fat and intra-muscular fat along with birth and carcass weight are used to resolve traits of interest when real carcass data are not collected. This could create some changes in carcass fat and marbling EPDs, the two traits reported out of this model. Also, the variance components used as part of the analysis procedure were re-estimated for all carcass traits and the correlation between scan intramuscular fat and carcass marbling decreased from 0.70 to 0.54. Finally, because of the change in the way the updated genetic evaluation is handling the genomic component there can be changes to some animals’ EPDs.
2d Weaning Weight, Ultrasound Rib-eye Area, Carcass Rib-eye Area and Carcass Weight
Weaning weight will be used as the correlated trait as birth weight is not adequate to solve for weight traits. However, yearling weight will no longer be included as a correlated trait with carcass weight and carcass rib-eye area to account for selection that may occur. Instead scan rib-eye area and weaning weight are included with any available carcass data to resolve the traits of interest. Carcass rib-eye area and carcass weight will be the only two traits reported out of this model. Analyses comparing the previous carcass trait EPDs to those coming from the new analyses are very high, but can be different for some animals, especially because of the change in the way the updated genetic evaluation is handling the genomic component.
2e. Calving Ease, Calving Ease Total Maternal, and Birth Weight
The updated genetic evaluation uses a random regression statistical procedure to calculate calving ease and maternal calving ease (total maternal calving ease). Birth weight is used as the correlated trait in this model but is not reported. The birth weight EPD that is reported comes from the Growth Traits analysis (2a.) discussed previously. This evaluation moves away from the threshold model previously used, in part because it required that observations that are the same within a contemporary group (e.g., herd) could not be used. This means that the new analysis can use all observations regardless of score as well as use the corresponding birth weight phenotype. Even though this evaluation uses only data from two-year old heifers, more observations will be used resulting in a more comprehensive calving ease evaluation than what is being performed today. Also, because of the change in the way the updated genetic evaluation is handling the genomic component there can be changes to some animals’ EPDs.
2f. Mature Cow Weight and Weaning Weight
Mature cow weight is analyzed with weaning weight but only mature cow weight will be reported out of this model. The weaning weight EPD that is published comes from the previously discussed Growth analysis (2a.). Even though the new MCW EPDs are highly correlated to the old analysis, small changes may be seen in part because yearling weight is no longer used as a correlated trait. Finally, because of the change in the way the updated genetic evaluation is handling the genomic component there could be changes in some animals’ EPDs.
2g. Sustained Cow Fertility
Sustained cow fertility is analyzed by itself, without a correlated trait, using random regression statistical model to predict female fertility/longevity. This trait predicts a female’s ability to stay in the herd through the age of twelve (ten calving’s after calving as a two-year-old heifer). The EPD is reported on a probability scale meaning that a higher EPD for a sire means his daughters are more likely to remain fertile and produce more calves throughout their lifetime. Because of the limited number of phenotypes collected that have a corresponding genotype, the genomic component is not included in this model.
2h. UDDER and TEAT
The udder and teat evaluation remains the same basic model as it is today with the exception that it includes the genomic information to increase the accuracy of prediction.
2i. Dry Matter Intake, Weaning Weight and Yearling Weight
Dry matter intake includes weaning and yearling weight as correlated traits but dry matter intake will be the only trait reported out of this model. This EPD predicts the daily consumption of feed. Because of the limited number of phenotypes collected that have a corresponding genotype, the genomic component is not included in this model.
3. Data Pruning Strategy
In the updated genetic evaluation, phenotypes only from progeny born after 2001, the advent of Whole Herd Total Performance Reporting (TPR), are used. The pedigree data includes at least three generations of pedigree from the observations (i.e., through great-grand-parents). This data cutoff was implemented to reduce the biases from the incomplete reporting of data that were submitted prior to Whole Herd TPR. The Whole Herd TPR program is based on a cow inventory system. The older data were collected only when a breeder chose to register a calf. Sires or dams that have had progeny born on both sides of Whole Herd TPR may see EPD and accuracy values change due to not including progeny born before 2001 in the genetic evaluation. Being able to have a genetic evaluation backed solely by Whole Herd TPR leverages the AHA’s commitment to our performance program and increases our reputation as a leader among breed organizations.
4. Accuracy Calculation
Because of the computing power of BOLT, the updated genetic evaluation will be able to more precisely calculate accuracy. Accuracy is quite possibly the hardest piece of a genetic evaluation to correctly compute and because of this complexity an approximation technique has been used to calculate accuracy by all breed organizations, including AHA. Because of the vastly improved computing methods a technique known as sampling allows the direct calculation of key variables used in calculating accuracy rather than approximating these variables. The previously used approximations overestimated the accuracy of EPDs especially for young animals. Accuracy of animals can change with the updated evaluation because the direct method does not contain the bias from the approximations.
5. Updated Economic Indexes
Both Dry Matter Intake (DMI) and Sustained Cow Fertility (SCF) will now be included in the AHA economic indexes along with other key Economically Relevant Traits (ERT’s) Carcass Weight (CW) and Mature Cow Weight (MCW). Adding these ERT’s into the economic indexes will provide a more robust and comprehensive selection tool for commercial producers to select Hereford bulls to be used on Angus based cows. DMI and CW will be included in all three AHA economic indexes to help predict the cost associated with feed inputs but to measure the end-product pounds that are critical for profit. SCF will replace scrotal circumference as the predictor of fertility and be a large contributor to both maternal indexes. Because of the inclusion of these key ERT’s animal index values may change. Watch for future Hereford World articles breaking down the key differences of the new index calculations.
The AHA would not have been able to complete this genetic evaluation overhaul without the guidance of several key scientists. First include Dr. Bruce Golden, Co-Founder, CEO and President of Theta Solutions, LLC, and Dr. Dorian Garrick, Co-Founder, CSO and CFO of Theta Solutions, LLC, that have developed the BOLT software and engineered the script writing and provided the genomic expertise for the updated genetic evaluation. Next Dr. Brad Crook and Dr. Shalanee Weerasinghe of Agriculture Business Research Institute (ABRI) who have orchestrated the genetic evaluation to run with Canadian Hereford and provided the technical support to ensure the evaluation is performing accurately. Also Dr. Mike MacNeil, Delta G, Dr. Matt Spangler, University of Nebraska, and Dr. Larry Kuehn, Meat Animal Research Center, have provided expertise and technical backing in the update of the new economic indexes. Finally, an advisory committee was assembled to review the revisions made to the updated genetic evaluation. This committee was charged with providing feedback to the proposed changes and ultimately giving the recommendation of approval to the AHA BODs. Advisory committee member are as follows; Joe Ellis, Jack Holden, Paul Bennett, Lee Haygood and Mitch Abrahamsen.