Blog: A Steak in Genomics

Help Support Steer Planet:

HerefordGuy

Well-known member
Joined
Apr 21, 2009
Messages
442
Location
Sturgeon, MO

knabe

Well-known member
Joined
Feb 7, 2007
Messages
13,639
Location
Hollister, CA
what would be the accuracy of their cattle with their own data versus adding database data?

what is the range of change of an epd over time of an animal and has this been plotted anywhere say for ext over 10 years.

probably ranking changes have an effect on numbers as well and how is this tracked over time so that people can see this. it seems a lot of analysis that could be simple to produce is not.

did their numbers go up or down with the addition of their data and what was the accuracy increase?

how do their numbers change with data from other producers using their bulls or other animals.
 

HerefordGuy

Well-known member
Joined
Apr 21, 2009
Messages
442
Location
Sturgeon, MO
Knabe- I thought I answered this post, but now I don't see it. Must have closed the browser before hitting Post. Let's try again.
knabe said:
what would be the accuracy of their cattle with their own data versus adding database data?
Not sure I understand this question. Are you asking about the difference between EPDs calculated on a single herd versus EPDs calculated nationally?

knabe said:
what is the range of change of an epd over time of an animal and has this been plotted anywhere say for ext over 10 years.

probably ranking changes have an effect on numbers as well and how is this tracked over time so that people can see this. it seems a lot of analysis that could be simple to produce is not.
I personally don't have any of this data. But, I think Sally Northcutt at AAA may track some of this for a small number of sires. I'll touch base with her.

knabe said:
did their numbers go up or down with the addition of their data and what was the accuracy increase?
When we add genomic tests to the EPD of a young animal the accuracy increases about 0.2, give or take. This is equivalent to about 20 progeny records. If the animal is older and already has an accuracy around 0.3, the increase is about 0.1.

knabe said:
how do their numbers change with data from other producers using their bulls or other animals.
Don't know. I assume it depends on the size and structure of the contemporary groups.
 

knabe

Well-known member
Joined
Feb 7, 2007
Messages
13,639
Location
Hollister, CA
HerefordGuy said:
Knabe- I thought I answered this post, but now I don't see it. Must have closed the browser before hitting Post. Let's try again.
knabe said:
what would be the accuracy of their cattle with their own data versus adding database data?
Not sure I understand this question. Are you asking about the difference between EPDs calculated on a single herd versus EPDs calculated nationally?


yes.  when new bulls start out with wider distribution, people sample lightly, which could make a bull less accurate than he really is, or even less real than he is, especially if used on a wide diversity of cows and limited access to best cows which again would skew data, especially when compared to within herd use with the same overall numbers, which may or may not happen depending upon lots of things. the whole thing is a little suspicious to me as i can't see how these things help small producers that have limited means to validate their epd's.  it's one reason people with 1000 or more cows have an advantage, they get better sampling on whatever range of cows they see fit, ie a changer population (worst cows), best cows population, diverse cow population.  all of these may have different effects on data and it would be interesting to just see how this data looks.  i don't think anyone really gets to see data. i guess i will just be a little skeptical until causal relationships can be found rather than just fence posts linked to long stretches of fence continue to be in vogue. yes, i understand these traits have markers with low % contribution and one needs "lots" of them leading companies to state that the average producer can't possible tabulate and introgress. instead, we still have to use numbers and have no idea if one is losing markers. sorry for the rambling.


on a side note, say there is 100 marbling markers and 100 tenderness markers, of those, what percentage work across breeds, what percentage work across what breeds and when will we be at a point where any across breed markers can be useful for anyone other than angus.


for instance, the top three bulls below i have semen on. continually, the maine association marbling numbers keep moving towards the average with few if any outliers.  these four bulls are outliers with a limited feeding dataset available, which sadly may not be available.  one of these bulls comes from a herd with a known sharp pencil but supposedly were at the top in an actual trial.  how could one possibly know if anything is real and seek to use this semen on something creative rather than just wasting it out of novelty?


http://www.maine-anjou.org/pagedisplay.php?id=63574


his sire, which people have semen on
http://www.maine-anjou.org/pagedisplay.php?id=10049


and an unrelated bull
http://www.maine-anjou.org/pagedisplay.php?id=340297


another unrelated bull with suspicious epd including his sire and then sub 0.2
http://www.maine-anjou.org/pagedisplay.php?id=278512


most of the rest of the breed is around 0.2 or less for marbling.  most of the breed through several rounds of epd updates has gravitated towards 0.2 with less variability each time.


now of course all the above could be just error and everything really is around 0.2 or less with the only bulls that excel at marbling are one's that excel at no marbling like cunia, the highest used bull in the breed.


http://www.maine-anjou.org/pagedisplay.php?id=257542
 

knabe

Well-known member
Joined
Feb 7, 2007
Messages
13,639
Location
Hollister, CA
Jared, did bif conference fees go down? i see it's 250. it seems like it was significantly more in the past. or just bad memory.
 
Top