Genomic predictions have as much value as the unpredictable and inconsistent runs thus far. Without any daughters, sires changed significantly and heifers were downgraded. Such inconsistency begs that question, which one is valid? If these are set genomic data being analyzed why does the model or criteria change and why did the formula change as well. Predictability presupposes some range of reason so given the unpredictability one can assume their may not be reason to rely on genomics for much at this point.

Just because genes may be present does not mean that they will be transmitted. There is a difference between finding genes in the pool and having those genes be expressed in an animal that is bred. The fact is that genes that are documented may not work at all in a mating. See article below.

BREEDING HOLSTEINS FOR GENETIC IMPROVEMENT

BY GREGORY S. WALZ 4-11-10

Breeding Holsteins for genetic improvement is not necessarily evaluated by looking at the resulting animal but more a reflection of the pool of genetics from which to breed from.

The philosophies of breeding include breeding according to type, according to TPI, according to aAa, according to Net Merit, according to cow families among others. Some even believe that breeding is the product of random luck. But these philosophies do not lead necessarily to an improvement in the genomic makeup of the animal. A more complex and innovative approach needs to be taken to make progress genetically.

In order to understand how genetic improvement can be made in a Holstein cow, one must first recognize that not all traits that are transmitted from generation to generation are the product of a dominant gene but rather may be the product of a recessive gene. This can be readily seen when we breed for a red Holstein. In order to breed for red, the primary way to do that is to have two red recessive genes at the genomic site. Since black is dominant, a single black gene and a single red gene in an animal expresses itself as black. Assuming that a genetic trait is transmitted only as a recessive, that trait may never appear unless it is done for a number of generations, and then only it may express itself as a matter of what appears to be luck.
But even this is a simple view of the breeding variants for red Holsteins because there is the red/black gene associated with Triple Threat and also the Telstar wild dominant red. These variants are further variations of red in Holsteins.

The next principles that need to be taken into consideration when trying to understand how to make genetic improvement in Holsteins are hybridization and line breeding. These two concepts seem to be divergent but are related. A basic assumption in the universe postulates that life forces involve both forces of attraction and dissociation. Assuming this is the case, a concrete example of this concept as it relates to breeding Holsteins can be seen by examining the forces attendant to a magnet. If you put two similar sides of a magnet together, they reject each other; if you put opposites of a magnet together, they have a strong attraction. Magnets also have the ability to associate in strings or sequences which suggests that multiple genes may be associated together when breeding for a trait in Holsteins. For instance, to breed for a good udder, a pattern of genes needs to be passed on to the next generation to succeed.

Applying these two principles of hybridization and line breeding to breeding cows for genomic improvement, the associative and dissociative characteristics of genes seem to be the basis for the belief that crossbreeding is effective at the genetic level because it appears effective at the animal level. It appears effective in an animal because the vigor of the animal seems to be the product of the intensity of the opposite genes. But the next generation is fraught with significant problems because there is no clear breeding direction any longer but only extremes that are ineffective animals for the breeder‘s purposes.

On the other hand, the lack of hybrid vigor may also be the reason for believing that line breeding is not a genetically sound approach. When there is a lack of hybrid vigor, we call this in breeding. But, when line breeding, it is important to understand that the genetic force to resist combining to related genes has a strong influence. The choice to seek an opposite is dominant when cross breeding because it has the greatest attraction, but the genetic choice to seek from slightly similar kinds with the least resistance is more complicated and not as predictable. With the magnet analogy, you would have different size magnets with different strengths and the genetic choice with the least resistance–the one with the most unlike genes–would be the genetic combination that occurs in a mating. Therefore, the effort to breed for genetic improvement in Holsteins is often not the product of opposites that attract but more likely the path with the least resistance in a genetic combination–the genetic combination with the least resistance. When this principle is applied while line breeding, the most similar genes likely do not become part of a mating but those genes that are the least similar do. So effectively line breeding likely results in similar genes from related animals not staying in the mating as a rule since through the process of natural selection compounding genes from related animals would have resulted in defective and inferior animals that would not have been able to survive or compete. This principle makes diversity the norm at the gene level and not uniformity. Therefore, having a clear direction in breeding has importance over time in developing an animal that can effectively achieve a breeder’s objectives from generation to generation by taking these principles into account when seeking to make genetic improvement.

There is a genetic reason for a gene combination or mating taking the path of least resistance, or the most unrelated, not the most related— this perpetuates the consistency in the breed. This makes the change in the Holstein breed take time and also makes certain that before a new genetic variant is introduced into the breed that the new variant has the capacity to thrive. On the other hand, if change were always to occur based on the genes that are the most similar, the genetic progress or change in the breed would proceed exponentially and breed consistency would be quickly lost. The breed would also have failed to continue to survive. Therefore, gene combinations are likely the product of choosing the path of least resistance in a mating more often than not or the genes most unlike each other. This process invites the most variation in the breed on the gene level so that the breed can readily adapt to the environment, yet also, on the animal level, makes the breed appear the most consistent over time because the oldest gene has the greatest probability of being the most different from the newest gene that the breeder is trying to develop in the Holstein animal. These principles promote the adaptability of the species by incorporating the greatest genetic variation in an animal and also promotes the perpetuation of the breed over time by utilizing what likely has been the core genetic pattern for generations of that breed. So the process of gene recombination or mating at the genetic level seeks to combine the most dissimilar to create an animal that appears the most consistent in the breed. Gene recombination, through the process of natural selection, produces the mean average of the animal’s genetic pool. The tension between the dissimilar genes makes each side of the genetic variant in the new animal tend toward the norm for the breed since the expression of the genes in the animal at a particular gene site is the product of both sides of the genetic combination.

So effectively being able to transmit a genetic trait is complicated. We need to recognize that the only reliable way to breed for genetic improvement in the Holstein breed is to utilize genetic patterns unique to a bull or a cow family. When correct breed patterns from a bull or cow family are used on an individual animal, change in that animal and at that genetic site is more likely to occur. In order to effect a change at a genetic site, the genetic pool must be sufficiently modified at that site so that when a new mating occurs, the desired characteristic can express itself. Patterns unique to a bull or cow family may be dominant in that animal and will more likely express themselves through the mating. In order to make a genetic change at a site, it is better to breed to the correct pattern and not breed to the extreme opposite since the likelihood that a positive mutational change will occur at a genetic site is low. The animal may appear correct, but the genetics will not transmit correctly when extremes are used. So to effect change more predictably when seeking genetic improvement you breed by correct patterns not by extremes since transmittable change is more likely then. Breeding to the extreme may create a good animal, but will not allow that good animal to effectively transmit its characteristics to the next generation.

Therefore, breeding Holsteins for genetic improvement requires that you breed to correct patterns and not breed to extremes. This breeding philosophy is valid if you evaluate the quality of the genetic pool to breed from and not just the resulting animal. A sound breeding philosophy then is to breed to the correct pattern and not to breed to the extreme for consistent genetic progress that will likely test out high genomically as well.

Good luck breeding cows. It is fun; it is challenging; it is science and art.

Gregory S. Walz

______________________________________________

Genomics from the US Jersey point of view

JERSEY GENETIC SUMMARY UPDATE

Comments from AJCA Executive Secretary Neal Smith ..

Spring has sprung—at last—and it comes with important developments in Jersey genetic evaluations.

Updates to the Jersey Performance Index formula were implemented with the April summary. Traits and their relative weights (with changes from JPI₀₆ noted in parentheses), are 42% PTA protein (+2%); 15% PTA fat (-5%); 12% Productive Life (no change); 6% Somatic Cell Score (+3%); 10% Daughter Pregnancy Rate (+3%); and 15% Functional Trait Index (no change). The updated formula puts 57% of emphasis on production, 19% on herd life, 14% on udder health and 10% on fertility.

The technical advances in genomics continue. For April, AIPL included more sources of information in genomic evaluations and also made adjustments for upward bias in genomic PTAs to increase reliability.

The staff at AIPL are doing amazing things with genomic technology, including finding ways to use genomic information from family members to calculate genomic PTAs for animals that have not themselves been genotyped. The newest method is called imputation, whereby the genotypes of a cow’s progeny (usually five or more) are examined in order to assess which genes came from the dam. The result is a genotype for the cow, not as precise as if she had been tested herself but far more accurate than a traditional PTA. The method is well-founded in the science and improves reliability of evaluations.

The second advancement deals with a problem that geneticists have recognized for some time. That is, the current system over-evaluates transmitting ability, particularly for individual cows, in part because there is never as much information about them as there is for bulls who have many daughters.

Practically speaking, this upward bias has not been a problem, because all animals have been evaluated using the same system. What mattered was not the exact numbers, but rather the rank of an animal or a group of animals compared to others. For example, think P9 heifers versus a group of P4s, or a young bull with Net Merit $400 compared to one with NM $250.

Still, over-estimating the cow PTAs is not desirable because of how it trickles down and affects PTAs of related animals. A well-known example of this is a young bull’s Parent Average. PA typically over-shoots what the bull’s proof will be after he is progeny tested. Why? Studies show that it’s because the dam’s contribution to his Parent Average is too high.

AIPL geneticists have studied this problem with the goal of increasing reliability of genomic PTAs. Their work shows that cows, young bulls and heifers that are genomically tested are over-evaluated by about 250 lbs. milk, 8 lbs. fat and 5 lbs. protein. To correct this upward bias, the evaluations of genomically tested animals have been adjusted downwards. This has had the effect of improving reliability and also impacting the evaluations of descendants.

What exactly do these advances mean? Genomic Predicted Transmitting Abilities (GPTAs) cannot be compared in any understandable way with traditional PTAs and PAs. Therefore, new reports have been created listing the top-ranking genomically evaluated cows and heifers. You will find these at http://greenbook.usjersey.com/CowsHeifers.aspx

Breeders have contacted us asking, “But why make the adjustment only to the genomically tested cows? Why wasn’t the bias adjusted out of all cows?” These are good questions, and we have taken them up with the staff at AIPL. It is an issue that we will continue to work on in the coming weeks and expect that there will be a satisfactory solution in the near future.

This entry was posted on Saturday, April 17th, 2010 at 9:05 am by kknutsen and is filed under Miscellaneous. Hits for this post:1253 You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.