librarian
Well-known member
The original issue on this topic, as I understood it, was the breeder was switching away from Shorthorns because he felt The association had not been aggressive enough in improving the market for commercial animals. Quality has never been the problem, just recognition.
Genomics are here to stay and, after studying on it, it seems companies like Igenity are somewhat under the gun to improve the accuracies of their tests across breeds.
The nature, scope and impact of genomic prediction in beef cattle in the United States
Dorian J Garrick
http://www.gsejournal.org/content/43/1/17
They need training populations to do this.
Because the genetic distance between Angus, Red Angus, shorthorn and Maine is short, the Results from Angus training populations should show high correlations to accuracy (as accuracy goes) for Shorthorn.
These tests are $40 per animal. If you are looking at 700 lb feeders, my math says that comes to almost 6 cents/lb.
I'm not great at math, so maybe that's not right. But if it is, it seems worth it to offer that information to buyers, especially if the animals are going straight to the buyer. Putting together 40 animals for a pen doesn't seem daunting. And putting together 10 pens doesn't sound daunting.
But we shouldn't even have to pay that $40, because the companies need the data. Here is where the Association can help, by making deals with the Genomics companies to use our feeders as training populations.
But here is the bigger thing:
They also need training populations of cross bred animals to assist is making predictions for purebred sires that work well in crossbred scenarios. So we could really accomplish something by getting into this kind of research.
We have what they need, so we just need to offer them numbers. Coming up with crossbred numbers should not be so hard. The thing is to make sure they are shorthorn sired to try to disprove the hypothesis that shorthorn sired cross bred animals out of Angus and red angus have better feed efficiency, quality and grade.
And that Shorthorn sired replacements out of Angus and Red Angus excel in maternal traits and stayability.
Sorry to put all this research up, but it we could fit right into this need for further study.
Genomic selection in admixed and crossbred populations1
A. Toosi, R. L. Fernando 2 and J. C. M. Dekkers
https://www.animalsciencepublications.org/publications/jas/articles/88/1/32
The final paragraphs:
A population that is a crossbred or an admixture of different breeds can be used as a training data set for GS and can provide reasonably accurate estimates of true BV of purebred selection candidates. This also implies that, with GS using high-density SNP markers, marker estimates obtained from crossbred populations can be used to select purebreds for crossbred performance, as suggested by Dekkers (2007), and examined by Ibanez-Escriche et al. (2009). Our results showed that in crossbred and admixed populations, haplotypes with strong LD are much shorter than in purebred populations. Thus, crossbred or admixed populations are more suitable for QTL fine mapping than purebred populations, provided marker density is sufficient.
Furthermore, because haplotype segments with strong LD in crossbred and admixed populations are narrower, markers in such segments are expected to have more consistent associations with QTL across the training and validation populations. Therefore, the decline of accuracy of GS over generations that has been observed in simulation studies (e.g., Habier et al., 2007) might be slower when admixed or crossbred populations are used for training than when purebred populations are used. By combining 2 pure breeds into a single training population, one can take advantage of a larger sample size for simultaneous estimation of marker effects and thus improve the accuracy of GS. In our simulation, when the size of the training population for the combined_AB training population was doubled, a 7% increase of the accuracy resulted (data not shown). In addition, by combining breeds into a single training population (vs. making certain crosses like an F1), a lot of time and effort can be saved. More importantly, there is a greater chance of segregation of breed-specific QTL in a multi-breed training population.
In the present study, while dealing with admixed populations, the population structure or additive genetic relationships were not explicitly modeled, which might be regarded as the standard method to limit the false discoveries due to population admixture in marker-phenotype association studies. Nevertheless, GS using high-density markers proved to be efficient enough to distinguish between true signals of association from spurious signals, at least under the idealized population structures that were used in the simulations. Whether or not this could provide an alternative methodology for association studies in populations with cryptic structures or extensive genealogical relationships requires further research.
Genomics are here to stay and, after studying on it, it seems companies like Igenity are somewhat under the gun to improve the accuracies of their tests across breeds.
The nature, scope and impact of genomic prediction in beef cattle in the United States
Dorian J Garrick
http://www.gsejournal.org/content/43/1/17
They need training populations to do this.
Because the genetic distance between Angus, Red Angus, shorthorn and Maine is short, the Results from Angus training populations should show high correlations to accuracy (as accuracy goes) for Shorthorn.
These tests are $40 per animal. If you are looking at 700 lb feeders, my math says that comes to almost 6 cents/lb.
I'm not great at math, so maybe that's not right. But if it is, it seems worth it to offer that information to buyers, especially if the animals are going straight to the buyer. Putting together 40 animals for a pen doesn't seem daunting. And putting together 10 pens doesn't sound daunting.
But we shouldn't even have to pay that $40, because the companies need the data. Here is where the Association can help, by making deals with the Genomics companies to use our feeders as training populations.
But here is the bigger thing:
They also need training populations of cross bred animals to assist is making predictions for purebred sires that work well in crossbred scenarios. So we could really accomplish something by getting into this kind of research.
We have what they need, so we just need to offer them numbers. Coming up with crossbred numbers should not be so hard. The thing is to make sure they are shorthorn sired to try to disprove the hypothesis that shorthorn sired cross bred animals out of Angus and red angus have better feed efficiency, quality and grade.
And that Shorthorn sired replacements out of Angus and Red Angus excel in maternal traits and stayability.
Sorry to put all this research up, but it we could fit right into this need for further study.
Genomic selection in admixed and crossbred populations1
A. Toosi, R. L. Fernando 2 and J. C. M. Dekkers
https://www.animalsciencepublications.org/publications/jas/articles/88/1/32
The final paragraphs:
A population that is a crossbred or an admixture of different breeds can be used as a training data set for GS and can provide reasonably accurate estimates of true BV of purebred selection candidates. This also implies that, with GS using high-density SNP markers, marker estimates obtained from crossbred populations can be used to select purebreds for crossbred performance, as suggested by Dekkers (2007), and examined by Ibanez-Escriche et al. (2009). Our results showed that in crossbred and admixed populations, haplotypes with strong LD are much shorter than in purebred populations. Thus, crossbred or admixed populations are more suitable for QTL fine mapping than purebred populations, provided marker density is sufficient.
Furthermore, because haplotype segments with strong LD in crossbred and admixed populations are narrower, markers in such segments are expected to have more consistent associations with QTL across the training and validation populations. Therefore, the decline of accuracy of GS over generations that has been observed in simulation studies (e.g., Habier et al., 2007) might be slower when admixed or crossbred populations are used for training than when purebred populations are used. By combining 2 pure breeds into a single training population, one can take advantage of a larger sample size for simultaneous estimation of marker effects and thus improve the accuracy of GS. In our simulation, when the size of the training population for the combined_AB training population was doubled, a 7% increase of the accuracy resulted (data not shown). In addition, by combining breeds into a single training population (vs. making certain crosses like an F1), a lot of time and effort can be saved. More importantly, there is a greater chance of segregation of breed-specific QTL in a multi-breed training population.
In the present study, while dealing with admixed populations, the population structure or additive genetic relationships were not explicitly modeled, which might be regarded as the standard method to limit the false discoveries due to population admixture in marker-phenotype association studies. Nevertheless, GS using high-density markers proved to be efficient enough to distinguish between true signals of association from spurious signals, at least under the idealized population structures that were used in the simulations. Whether or not this could provide an alternative methodology for association studies in populations with cryptic structures or extensive genealogical relationships requires further research.
