You Would be Crazy not to Test!

Things that come up in the daily operation of a ranch.
User avatar
Mike
Rancher
Rancher
Posts: 28542
Joined: Thu Feb 10, 2005 4:05 pm
Location: Montgomery, Al

Post by Mike » Sun Jun 09, 2013 10:09 pm

knabe wrote:
Mike wrote: Genetic tests for complex traits are likely to require hundreds or even thousands of
markers to effectively track all of the genes influencing complex traits
which says find a different way.

if a marker has 0.1% impact, it's useless and so is the method to track it.
Yep. There is probably no way to identify all the genetic markers associated with Birth Weight or Weaning Weight for example. It will take a loooong time for DNA markers to affect EPD's in those traits.

But on the other hand it's fairly simple to tell whether an animal is Homozygous or Heterozygous for horns, coat color, etc. via DNA.

Plus, unless you have a bull whose semen is shipped to most parts of the country and bred to several hundred dams, his EPD accuracy values will never get high enough to be reasonably predictable with measurable data.


I'd rather have a bottle in front of me than a frontal lobotomy.

HerefordGuy
Member
Member
Posts: 132
Joined: Fri May 24, 2013 5:09 pm
Location: Columbia, Missouri

Post by HerefordGuy » Mon Jun 10, 2013 7:48 am

knabe wrote:
Mike wrote: Genetic tests for complex traits are likely to require hundreds or even thousands of
markers to effectively track all of the genes influencing complex traits
which says find a different way.

if a marker has 0.1% impact, it's useless and so is the method to track it.
knabe-
This is not true.
Using single gene marker tests and using genomic selection are completely different paradigms. Gene markers may work well in plant breeding. They do not work well in livestock quantitative traits. Genomic selection works well in livestock, because it takes the region with a 0.1% impact and uses it and the rest of the genome to create predictions.
http://steakgenomics.blogspot.com/2012/ ... ction.html

knabe
Member
Member
Posts: 123
Joined: Wed Jun 27, 2007 3:39 pm
Location: CA

Post by knabe » Mon Jun 10, 2013 9:10 pm

.35 x 0.001 = 0.035% of total variability can not be measured.

If it is, rank 100 bulls on a continuum with 5% difference in marbling in various states of homo/het status, detail the markers, progeny and rank each marker' contribution range with confidence intervals etc, id a bunch of calves, place them all in the same feeder trial and predict marbling on a 1% basis and rank them all with less than 5% out of order.

HerefordGuy
Member
Member
Posts: 132
Joined: Fri May 24, 2013 5:09 pm
Location: Columbia, Missouri

Post by HerefordGuy » Mon Jun 10, 2013 10:29 pm

knabe wrote:.35 x 0.001 = 0.035% of total variability can not be measured.

If it is, rank 100 bulls on a continuum with 5% difference in marbling in various states of homo/het status, detail the markers, progeny and rank each marker' contribution range with confidence intervals etc, id a bunch of calves, place them all in the same feeder trial and predict marbling on a 1% basis and rank them all with less than 5% out of order.
because it takes the region with a 0.1% impact and uses it and the rest of the genome to create predictions. Genomic prediction is radically different from previous genetic tests, because it uses markers throughout the entire genome.

Genomic selection (a.k.a. a genomic-enhanced EPD) works. The rate of genetic change increased when the dairy industry implemented genomic selection. See this PDF. http://www.cdn.ca/document.php?id=281

Badlands
Member
Member
Posts: 437
Joined: Tue Apr 11, 2006 10:53 am
Location: East Idaho/Western MT

Post by Badlands » Tue Jun 11, 2013 7:13 am

HerefordGuy,

You are attacking the genomics side with the same youthful exuberance shown by your predecessors back to the late 70's. A couple of your colleagues at Mizzou are entirely capable of selling snowcones to the Eskimos mentioned earlier in this string.

As stated by one of my professors, "The genomics revolution has been "just a few years around the corner" since I started my career in 1975". He is now Emeritus, LOL.

So, are we now "just a few years around the corner"?

With a PhD, you know as well as I do that the beef industry is a different beast than any of the other animal industries. You also know about the emerging field of epigenetics, and that some things may not be entirely what they appear to be.


Tell me about the possibilities of cattle being discriminated against because of their genomic profile, but having those genes down-regulated because of their fetal environment. Do you think it is a possibility that commercial cattle could be discriminated against because of their profile, but they might actually be good performers because the "bad" genes have been down-regulated?

I am concerned that the "full speed ahead" approach in the application of genomics may be detrimental to a large portion of the western US producers. What if the "range cattle" genes reduce "feedlot" performance? What if we ultimately just end up with "feedlot" genes, that are nothing more than growth genes, and the level of growth genes the buyers want will make cows too big for arid environments? Please address this with any information you have.


You said, "I know most of these bulls are purchased for their appeal in the show ring." Well, right there, they have separated themselves from reality. Same as the other bull you mentioned.

Have you seen the genomic profile of the bull from years ago you mentioned in your blog? It would be interesting to compare his enhanced interim EPD with his original interim EPD. You have to remember, part of the reason he was purchased was because he was "pretty" compared to other bulls from the ranch. You also have to know that despite his purchase price, he was paid for in short order because of high certificate prices. In other words, he made his owners money, regardless of whether he was the "right" bull. Those bulls will continue to be around regardless of their genomic profiles. It is partially where the show ring screws things up in the industry. Same thing happens in the other animal industries. The pretty show animals are often loaded with genomics that affect eating quality and animal welfare.



Badlands

HerefordGuy
Member
Member
Posts: 132
Joined: Fri May 24, 2013 5:09 pm
Location: Columbia, Missouri

Post by HerefordGuy » Tue Jun 11, 2013 12:54 pm

Badlands,
You inspired a blog post! http://steakgenomics.blogspot.com/2013/ ... cycle.html
If I have used hyperbole, it has been a) by accident or b) pushing back against the disillusionment. Are genomic technologies the panacea http://dictionary.reference.com/browse/panacea to all the ails the beef industry? No. But, they are a useful tool.

I got my BS degree from NMSU. NMSU managed a herd of Brangus and a herd of Angus in the Chihuahuan Desert. They selected AI bulls that had superior calving ease and that were moderate sized. The tools they used to do this were EPDs! We haven't changed the tool, we are adding additional information and data into the tool of EPDs. This data allows us to gather information on a wider set of animals and produce accurate EPDs earlier in the animals life. Cattlemen are still responsible for making correct breeding decisions for the environment and production system. Genomic-enhanced EPDs allow them to use natural service sires and young AI sires with more confidence.

The bull I mentioned in my blog is almost 10 years old. He was originally marketed as a performance bull with excellent calving ease. Unfortunately, the calving ease didn't pan out. But, if genomic-enhanced EPDs would have been available, it would have been as if the bull had 30 progeny with calving ease records when he was sold as a yearling.

Epigenetics will affect the management of cattle, but I don't think it will affect the selection of cattle. Epigenetics is currently handled in our genetic predictions as a contemporary group effect or as an environmental effect.

PATB
Member
Member
Posts: 571
Joined: Fri Apr 10, 2009 5:55 am
Location: Turner, Maine

Post by PATB » Tue Jun 11, 2013 1:08 pm

HerefordGuy what is the difference between zoetis 50k and geneseek ggp-hd test? Is there an accuaracy difference between the 2 test?

User avatar
gcreekrch
Rancher
Rancher
Posts: 11338
Joined: Thu Feb 21, 2008 11:35 pm
Location: west chilcotin bc
Has thanked: 2 times
Been thanked: 2 times

Post by gcreekrch » Tue Jun 11, 2013 1:37 pm

Me thinks some folks have too much time on their hands.

There never has been a perfect cow herd and there never will be! :wink:


HerefordGuy, do you own cows? Another expert in his own field gave a seminar several years back about developing a perfect herd of cattle that a friend of mine attended.

After listening to the long presentation, my friend asked where the speaker lived.

The speaker asked why he wanted to know.

My friend said he wanted to visit and look at this perfect herd.

The speaker replied, "I don't own any cattle........." :roll:

As most here already know, I'm not the most politically correct member. :D
Don't tell people your problems, half of em' don't care and the other half are glad you got em' We can all run the neighbors better'n our own

HerefordGuy
Member
Member
Posts: 132
Joined: Fri May 24, 2013 5:09 pm
Location: Columbia, Missouri

Post by HerefordGuy » Tue Jun 11, 2013 1:52 pm

PATB
Great question!
As you know, both Zoetis and GeneSEEK have updated their genomic tests.
Zoetis still uses about 50,000 SNPs, but have now trained their data set with more animal records and genotypes.
The genetic correlations for the Zoetis test range from 0.38 to 0.73. http://www.angusjournal.com/ArticlePDF/ ... 003.13.pdf

GeneSeek now use about 80,000 SNPs. The genetic correlations for the GeneSeek test range from 0.60 to 0.70 for most traits, except milk which is at 0.4 and calving ease direct which is at 0.34. http://www.angus.org/AGI/GenomicEnhancedEPDsMay2013.pdf

So, the accuracy differences appear to be pretty small.

Both tests cost $75. If you add an AM, NH, or CA test to the Zoetis test, you pay $23.00. If you add an AM, NH, or CA test to the GeneSeek test, you pay $8.00.
https://www.angusonline.org/AGI/AgiDnaPricing.aspx

HerefordGuy
Member
Member
Posts: 132
Joined: Fri May 24, 2013 5:09 pm
Location: Columbia, Missouri

Post by HerefordGuy » Tue Jun 11, 2013 2:06 pm

gcreekrch-
I live in Columbia, MO. I hope to close on a farm house outside Sturgeon, MO with 12 acres in the next few weeks. After I own that small farm I hope to purchase some cattle. Until recently, I was a poor student trying to support a family. I am at least the 4th generation in my family to raise seedstock and commercial cattle, if you count the cattle I owned in my youth. I hope that is enough personal information for you! :D

I'm not talking about perfect cattle. I'm talking about cattle that are better and more effective than the previous generation. I'm talking about making a better decision when we purchase our next bred heifer or clean-up bull. I'm talking about shortening the generation interval so beef breeds can make genetic changes more rapidly.

Badlands
Member
Member
Posts: 437
Joined: Tue Apr 11, 2006 10:53 am
Location: East Idaho/Western MT

Post by Badlands » Tue Jun 11, 2013 2:39 pm

HerefordGuy;

so beef breeds can make genetic changes more rapidly.

Which trait is so "out of whack" right now that genetic change must be rapid?

After nearly 50 years of performance selection, what is wrong with our cattle?



Badlands

User avatar
gcreekrch
Rancher
Rancher
Posts: 11338
Joined: Thu Feb 21, 2008 11:35 pm
Location: west chilcotin bc
Has thanked: 2 times
Been thanked: 2 times

Post by gcreekrch » Tue Jun 11, 2013 3:25 pm

HerefordGuy wrote:gcreekrch-
I live in Columbia, MO. I hope to close on a farm house outside Sturgeon, MO with 12 acres in the next few weeks. After I own that small farm I hope to purchase some cattle. Until recently, I was a poor student trying to support a family. I am at least the 4th generation in my family to raise seedstock and commercial cattle, if you count the cattle I owned in my youth. I hope that is enough personal information for you! :D

I'm not talking about perfect cattle. I'm talking about cattle that are better and more effective than the previous generation. I'm talking about making a better decision when we purchase our next bred heifer or clean-up bull. I'm talking about shortening the generation interval so beef breeds can make genetic changes more rapidly.
Thanks, my best wishes to your endeavors.

I started with nothing and still have most of it left although we have added a few acres and cattle over the last 33 years.

I have purchased a lot of our herdsires from one large seedstock producer since 1993. The rest we have raised from our own commercial herd. None of these bulls ever see a show ring and I know their birth, weaning and performance weights along with their birthday. I also know what the "type" is of the herd they come from. Nothing more concerns me.

We sell a few loads of calves every year at premium prices to repeat buyers that claim our cattle perform well and don't get sick. Our ideal is to produce moderate, thick, broody females that need little intervention from us other than nutrition for replacements and let the steer end look after itself. So far it's not working too bad.

I personally don't see the need to change the type of cattle that work here. Others sometime like to follow the lastest, greatest propoganda in a mad rush to keep up with "progress". The 50's and the 70's show how well that worked for most other than those who started the fads. :wink:

When you produce a cow that will live here on snowballs and promises while bringing home a good calf every fall you will have a good chance of getting my interest. Until then I'll just wait and watch. :D

You will find Tom Lasiter's quote about keeping it simple has a lot of merit as time passes. :D
Don't tell people your problems, half of em' don't care and the other half are glad you got em' We can all run the neighbors better'n our own

Post Reply