Great post at Border Wars.
Also check out the post debunking some dubious scholarship on Nova Scotia duck-tolling retriever COI’s.
October 11, 2011 by SWestfall3
Great post at Border Wars.
Also check out the post debunking some dubious scholarship on Nova Scotia duck-tolling retriever COI’s.
Posted in dog breeding | Tagged Border Wars, Coefficient of Inbreeding, Nova Scotia Duck-Tolling Retriever | 8 Comments
Comments are closed.
Something that the writer fails to take into account is that (a) the Coefficient of Inbreeding is only an *estimate of probability” — it may or may not reflect the actual genetic influence of those ancestors. Ch Whoozis may appear several times in a 10 generation pedigree: it is *probable* that his influence is stronger than Noname who appears but once– but it is also possible that Whoozis’s genetic contirbution has been so diluted, or lost altogether, that he actually is of no consequence whatever.
(b) With each succeeding generation from an individual, that individual bequeaths only one half of its genes to each offspring, and so on in turn for succeeding generations. And the “half” (or quarter, or eighth…) that is passed on in each generation, witll vary between offspring. It is entirely possible (not probable, but possible) for two siblings to each have inherited completely different genes from the parents. COI says their inheritance is identical: as any experienced animal breeder knows, even full brothers, or full sisters, may produce quite differently. Short of sequencing the entire genome of each (and their partners), there is no way of really knowing.
The reassortment of genes is one of Mother Nature’s ways of helping to maintain genetic diversity in a population, And keeping breeders from becoming complacent.
M.R.S. – And it’s POSSIBLE that you’re really a chimpanzee banging away randomly on a keyboard tucked inside a secret soviet base hidden beneath a White Castle restaurant on the outskirts of St. Louis as part of a sleeper plot to make Americans stupid by spreading misinformation on dog blogs.
Genetic probability is not like guesstimating the weather on one day: 20% chance of rain tomorrow is entirely different than 20% inbred. The confidence one would have in the first number would be very low, as it’s a one time probability based upon a very imprecise model. The confidence in the later calculation is very high given that biology and mathematics work as expected and we’re dealing with a very large number of genes.
This brings into effect the Law of Large Numbers. I’ll let you educate yourself on the issue.
http://en.wikipedia.org/wiki/Law_of_large_numbers
When you say things like “two siblings to have each inherited completely different genes from the parents” you forget to say that it’s also equally likely to that outcome that they are identical twins through random chance instead of biological means like an egg splitting, etc. AND that both of those outcomes are entirely less probable than winning the Powerball 5 times in a row.
How often do you spend each day doubting the obvious and substituting the POSSIBILITY of the ridiculous? When you make your morning coffee do you suspect that your coffee maker had been secretly stolen during the night only to be replaced mere minutes before you woke up by the burglar who suddenly became remorseful and decided to return it without leaving a trace?
Do you question if you had brain cancer last month only to have it go into spontaneous remission and disappear?
Do you think the Government is listening in on your phone calls?
Do you set aside time each day for the Publisher’s Clearing House van to pull up and present you with a giant check?
COI never says anything about inheritance being identical! All the puppies in any given single sire litter will have the same COI, but the calculation tells you nothing about WHAT genes are being doubled up! Of course there will be differences, no one said differently.
Your observations have nothing to do with the content of my posts.
The differences between siblings is really irrelevant as you can see clearly from the Toller pedigree that no one was breeding every sibling from a litter to maintain genetic diversity. In fact, the chart clearly shows that popular sires often created a popular sire son who created a popular sire son.
That third generation popular sire can’t suddenly find new “reassorted” genes from his father or grandfather. He got a finite set of genes from dad and dad got a finite set of genes from the original popular sire and those don’t change.
As for natural re-assortment, COI takes that into account just fine.
“It is entirely possible (not probable, but possible) for two siblings to each have inherited completely different genes from the parents.”
If the parents are closely related, as in a high COI litter in a breed with few founders, this is *extremely* unlikely.
It’s actually impossible. Any gene that is homozygous in the parent will appear at least once in all the children.
The number of alleles in any breed is finite and in many breeds actually very limited. This is why two Border Collies don’t suddenly produce a Pomeranian, or perhaps a Chimpanzee. They have a common ancestor, you know.
I see breeders make stupid comments like M.R.S. and they always forget to include the alternative. When faced with a fact like “Tollers have a COI of 35%” they say “but but but it could be 10%” … well, the odds of that are the same as it being 60%. Chew on that.
They always forget the upper limit when they want to lie to themselves with their teaspoon full of math knowledge.
I don’t do the maths, you know. Numbers confuse me. When I am calculating nutrients for my dog diets, there is frequently screaming.
In any case, *how the puppies look* is not necessarily a way to tell *which genes* were inherited from which parent. The entire litter that Enki came from looks like the dam. So did all the pups in her first litter. Both litters complete outcrosses. This doesn’t mean that they didn’t get any ‘head shape’ genes from the sire. It means that the genes they inherited from the dam were dominant over those from the sire. The dam is of Iranian origin, a gen 2, and the sire of Enki’s litter is Western, an outcross. The sire of her first litter was Iranian, also gen 2. You’d think there’d have been more variety but she put a clear stamp on her pups.
Not to mention that the genes with regulate appearance are only a small part of the genetic package.
Just for shits and giggles, Enki’s COI for the entire known pedigree is 1.27% HIs half-sister Minna’s is 0%, almost her entire pedigree is recent COO dogs. These are Salukis, btw, which have an extremely variable appearance anyways. Yet they look almost exactly alike, and just like their mother.
I should also note that in each and every case where we have an empirically measured level of homozygosity — when we use DNA analysis to look at how inbred an individual is — it’s always been higher than the calculated COI. There have been no “omfg look at this, the COI was way too high!” moments in genome science, ever.
It’s obvious why: Founders aren’t all unrelated and heterozygous for any breed.
So remember, COI is a calculation that depends on accurate information, but when the information is accurate and plentiful it’s extremely highly correlated with what we find via genome analysis. AND the limits of the calculation almost always result in it UNDERESTIMATING the level of homozygosity.
I’ll clarify one statement in my response: in the second paragraph, it would have been more clear to say “COI says that the percentage of inheritance from each ancestor is identical in quantity” although not necessarily in identical genes.
And yes, certainly, there are vastly different degrees of probability. Which contributes to making things all the more “interesting” in this old world.
Do chimpanzees still use typewriters? …thought they’d at least have upgraded to an iMac….