Genes Show Limited Value in Predicting Diseases - Page 2

x CEL ENT TO x CEL ENT iS NOT A gAR EN TEEEEE ,wHERE DID YA GET THAT FROM luv me dogs. at bEST IT IS STILL A crap SHOOT ,,,,,,,,,

Notice the majority of matings conducted (~86%M) are Good X Fair or better.  That's roughly the total % of non-dysplastic dogs.  If Good X Fair or worse matings were avoided, the overall breed condition would dramatically improve.  I'm so sick of "crapshoot" analogies, regarding hips, temperament, etc.  The odds can be worked either less OR MORE in your favor with anything.

okkkkkkkkkkk WHEN 2  DOGSSSS OUT OF 1000     TESTED PER MATINGS  .uhaaaaaaaaaa BELIEVE THE DATA ......... duhhhhhhhhhhhh only THE GOOD ARE ON FILE....looks LlkES THE TRUE TALE  IS NOT REPORTED ...How many times have you seem one dog out of a litter OFA registard .......... LOTSSSSSSSS .What happened to the rest oF the litter DATA >>>>>>>>???????????    

Not a bad point.  ;-)

No, not a bad point. To me, though, it suggests that many dogs aren't OFA tested simply because they are pets and are not used for breeding. True, some are not listed in OFA data because they are dysplastic and they unfortunately are used for breeding, but we have no way to tell what percent that would be.  Daryl Ehret's charts clearly show that breeding excellent to excellent produces very good results and that as the ratings of the parents decrease, the puppy's chances of being dysplastic increase.  So to limit your chances of getting a dysplastic pup, choose parents who are OFA good or excellent.  Great charts!

As much as I respect Daryl and his understanding of genetics and ability to manipulate data, I think he would be the first to say that in any problem, put in garbage data, and get garbage data out. The issue of any type of health problem relating to the GSD is a moot point at this time in the history of the breed because the data is lacking and so incomplete. It is an incredibly small segment of people breeding any kind of dogs who do any kind of health checks at all, let alone OFA or SV hip and elbow checks. So to make any suppositions based on such small numbers is likely invalid.

There is a second issue that we don't often discuss and it comes back to what one looks for in the "exemplary" specimen of the GSD breed. Because we are the manipulators, and so often not the animals themselves, we may, in fact, be doing much more than simply creating bottlenecks in genetic diversity. When one selects for any characteristic or trait, there can be unexpected consequences. In fact, to eliminate certain dogs from a gene pool because of certain traits mightnbspnbspnotnbspeven be a great idea.

Especially given the fact that so few characteristics or traits we look for are related to a number of unidentified genes, rather than simple autosomal dominant or recessive genes, or X linked traits, for example, means we will cannot even remotely manipulate the genetic heritage of the breed in a responsible fashion (assuming we could agree on what that is - and I am certain we could not agree). Perhaps it is a good thing that we cannot just yet do more manipulation. Just think, for example, what if the rage became to select for a certain genetic trait (of any kind), and in doing so, we further eliminated the working ability of the dogs so that they lost, for example, some of the drives we all want and talk about?

In everything you consider, there are incomplete datasets, and that’s why it’s necessary to put each use into their proper context.  When our human forefathers endeavored their many approaches to crack the code for defying the laws of gravity for example, which are nearly universal, there then emerged a field of study (aerodynamics) with a new set of troubles, which were inherently at the mercy of a new set of laws and dynamics unique unto itself (i.e., turbulence, payload, and consequently traffic control, terrorism).

Nothing is so totally “random” as to be equated as a “crapshoot”.  Just as the data for Excellent X Excellent offspring could be admonished for being incomplete, so fairly are the Dysplastic X Dysplastic.  You then may draw a relatively more complete picture, by considering the degree of variance between the two, as I have done with the SV rated progeny.  When there is a broader variance between the portion of progeny tested, with a larger portion rated “a1”, then manipulation of the data becomes more suspect.  Then also, there are considerations to consider that rating systems are rather different, and surveyed at different ages.  The 1 year SV hip rating may not be as “telling” as a 2 year OFA rating.

Of course nothing is guaranteed, except the fact that somebody will come along to point out the fact that nothing is guaranteed, this perhaps for their need to oversimplify and avoid comprehension of the finer details.  That gives cause for things to appear as if they’re random.  There are exceptions to every rule, infractions of every known law of science, and the best that you can hope for, is to reduce the degree of variance which invariably occurs on multiple levels, and miniscule amounts, and fewer frequencies, all by narrowing the context, so to become more comprehensible and less variable within the limits of our human perception.  But nothing is free of influence, and the tiniest factors can snowball into much larger effects, gone unnoticed in their initial stages.

Time goes fast when you’re in a coma, distance traveled swiftly in a jet, a two ton weight on your shoulders won’t feel much different than a one ton weight, and paying attention to details can pay off in the long term.  One man’s garbage may be another man’s treasure, so garbage is sort or relative, and can be recycled to be of some sort of use.

Good points Daryl. But even so, having more data before drawing conclusions would be best. And no, I agree, "crap shoot" is not correct either.

The larger dataset is almost always better in statistics, since the conclusions drawn from it will be more representative of reality. If I toss a brand new penny 20 times, calling heads each time, and get 18 tails and 2 heads, that doesn't mean I don't still have an equal chance of getting heads next time. If I toss the penny 1000 times, the numbers will approach half heads and half tails in this simple binary system.

With genetics, we don't even know the complete dog genome and how the various genes interact yet. We know bits and pieces for various coats, colors, and diseases. This is good.  More information is better, but you often have to go with what you've got. Daryl has one of the better systems that I've seen for selecting health and working ability. Many breeders simply rely on their knowledge of the parents' pedigrees and the  structure and performance abilities of the dogs in those pedigrees. This method depends on a breeder's knowledge, judgment and intuition. Sometimes it works, sometimes it doesn't. Neither of these methods are a crapshoot. But all we can say with either of them is that you increase your odds of getting the pups you want. I suspect but don't know that Daryl's method will give him more of the pups he wants more often than the other method would.

Actually, the canine genome map already exists. There are some fascinating and serious research projects going on world wide that eventually will hopefully lead to being able to treat diseases.