Genomics has become one of the most powerful tools in the cattle industry over the past decade, but the technology is often misunderstood. Genomics is a major part of nearly all seedstock genetic evaluations – i.e., expected progeny difference (EPD) calculations – but it is also beginning to see more use in commercial settings. My goal here is to help dispel some myths and confusion around the use of genomics.
Genomic tests are not the same thing as ‘gene tests’
Much of the industry’s first exposure to genomics was in the form of single-gene tests for carcass traits. These tests were designed to identify an animal’s genotype for a handful of large-effect DNA variants involved in marbling and tenderness. These single-gene tests were generated to identify mutations believed to have an outsized effect on traits of interest. These associations were highly dependent on the population of animals being tested, and many failed to be validated by follow-up analysis.
Contemporary genomic tests are very different in design compared to these early single-gene versions. Current tests are less interested in genotyping known mutations that affect specific traits. Instead, they aim to characterize the landscape of an animal’s 3,000,000,000-plus DNA bases. Genomic tests take advantage of the fact that DNA is inherited in large chunks called haplotypes. As such, a single marker is sufficient to characterize the likely state of large areas of the genome. Current tests use between 30,000 and 100,000 evenly spaced markers in categorizing the entirety of the cattle genome, rather than a handful of places.
In genetic evaluations, genomics increases EPD accuracy
EPDs are our best statistical guesses of an animal’s actual genetic merit. The ultimate goal of EPD calculations is to help identify which DNA, good or bad, an animal inherited from each of its parents. The random sampling of genetics that an animal receives from each parent can lead to considerable differences in genetic merit between siblings. EPDs use information on relatives and progeny to understand whether an animal received good or bad samples of parental genes. As the information supporting an EPD increases (i.e., progeny records, pedigree-relative progeny records), our confidence that the prediction represents the animal’s actual genetic merit also increases.
We usually choose from bulls with low or no progeny information. This is true of commercial herds purchasing young bulls and seedstock operations that use the newest generation’s genetics to drive progress. Historically, this lack of progeny information in EPDs meant low accuracy in young animals. With the addition of genomics, we can directly identify which genetics an animal inherited from its parents rather than relying on progeny information to resolve these differences. The result is EPDs that are more accurate for unproven animals. Depending on the trait, accuracy increases from a genomic test can be the same that 10 to 30 calves would generate.
EPDs with genomics are always better than EPDs without genomics
It is essential to remember that genomically enhanced (GE) EPDs appear and can be interpreted exactly as nongenomic EPDs. A genomic test does not change the traits that receive predictions. It only increases the accuracy of the reported predictions. This helps us increase the confidence that a prediction represents the bull’s actual genetics that it can deliver to our herd. Bulls sold without genomics tests are inherently riskier than bulls with GE-EPDs because their EPDs and indexes are less informative about their actual genetic potential.
EPDs can experience substantial change with the addition of genomics. These changes are not statistical noise but a better representation of the random sampling of parent genetics that an animal received. This means that whether or not the genomic test increased or decreased the value for a trait, the GE-EPD is a better representation of the animal’s actual genetics than the nonenhanced EPD. An increase, decrease or no change in the EPD itself may accompany this increase in accuracy. For every animal whose GE-EPD moves in a positive direction, we’d expect to see another shift the other way.
Using genomics without phenotype collection is counterproductive
There is a misconception that using genomics relieves us of the need to collect phenotypes. Nothing could be further from the truth! All EPDs, conventional or genomically enhanced, rely on the continued collection of phenotypes in the population. Our genetic prediction models must be trained on new data from new animals, or their quality will quickly erode. Even GE-EPDs receive a boost in accuracy when an animal’s actual phenotype is added to the evaluation.
Commercial genomic tests are not the same as GE-EPDs
One of the hottest topics in recent years has been the increased use of genomic tests in commercial operations. This has been driven largely by the decrease in the cost of genotyping. Commercial tests, while helpful in selecting heifers and marketing feeder calves, differ from how genomics is used to calculate GE-EPDs. Notably, no pedigree or phenotypic information is used to augment these genetic predictions. While the genotype alone is sufficient to generate a reliable prediction, they are significantly less accurate than true GE-EPDs from a National Cattle Evaluation (NCE). Further, the quality of these predictions relies on the tested animal’s breed(s) being adequately represented in the test’s training population.
Genomics helps reduce risk
The bull-buying process is one of the most important and riskiest things a beef herd does. Genomics is a tool to help reduce some of this risk when purchasing or using young bulls. Integrating genomics allows us to use EPDs with greater confidence that they represent an animal’s genetics. As tests have declined in cost, there are very few excuses for purchasing bulls without GE-EPDs.
