puppies tool

[buidelwolf]

The problem of measuring the COI on only a limited number of generations (even on 8 generations), is that most of the large inbreeding effect during the early years, which in particular Rep z Pohranièní stráze has brought, is simply being disregarded. In genetics, the most reliable figures regarding inbreeding are obtained if one measures the COI on as many generations as possible. In that respect it is regrettable that the COI measurement on ALL generations (temporarily?), is no longer possible on wolfdog.

Elf, being a fan of your statistic tool, does it ever get that possibility in the future?

Regards

[elf]

It would require some optimization but unfortunately I have not much time lately for updates. But I think the best would be to plug the database of this site to a pedigree software (like Pedserve), this would unload some work and provide many features.

[yukidomari]

Originally Posted by buidelwolf

The problem of measuring the COI on only a limited number of generations (even on 8 generations), is that most of the large inbreeding effect during the early years, which in particular Rep z Pohranièní stráze has brought, is simply being disregarded.

Yes, that is true, but some of the CW dogs I checked don't have compete pedigrees even in the 8th generation out. I have been reading about COI recently and came across this article. It doesn't talk about CW specifically but rather Australian Shepherds (some lines also don't have complete 10 generation pedigrees yet, like some CWs)..

Here is an excerpt:

"...The usual 3-5 generation pedigree won't give sufficient information for a useful calculation. For Australian Shepherds the author has found 10 generations to be the best indicator. Few Aussies have a complete 10-generation pedigree, so this will give you pretty much all there is to know. While some individual lines of descent may go back 20-30 generations, running the calculation for more than ten does not result in any significant change in the result. Running it for fewer than 10 frequently causes the number to drop, giving an incorrectly optimistic result...

Australian Shepherds are blessed with around 300 founders, so our background inbreeding is very low. However, our breed like most others has experienced historic inbreeding stemming from the choices made by recent generations of breeders who have used one sire more than others or frequently sought the output of a particular kennel. ...Even in working lines there are early sires and kennels whose names occur with considerable frequency."

I don't know if this is because Aussies have a larger foundation group (300), but if pedigrees for CW start to fray at around 8-9 generations back for some dogs, I think that an 8-10 generation COI is probably pretty accurate.

Still, I would like to be able to have the capability for all generations too!

http://www.ashgi.org/articles/breeding_incest.htm

ETA: Oh, wait.. I just reviewed some pedigrees again and the blanks in the 8th generations are those of the founder wolves. So then the question is.. how far back do the known pedigrees of the founding GSDs go?

[buidelwolf]

Hello Yukidomari,

Very interesting post!

but some of the CW dogs I checked don't have compete pedigrees even in the 8th generation out

All Cvs have a complete pedigree at least to the base. I.e. up to the four wolves and the german shepherds they are crossed with; in some cases, even a limited number of ancestors of the founding German shepherds are included in the family tree. By measuring the COI on all generations I meant all the generations upon this base of the breeding. That is/was the way one could measure the COI on "all generations" with Wolfdog.org until recently.

If you measure the COI of old Cvs (or deceased) over 8 generations (or more), it can happen that you do indeed encounter a number of empty holes in the pedigree, but in any case it is complete upon the base of the breeding project, with exception of the "mixes".

I understand that Wolfdog is busy with adding the pedigrees of German Shepherds all the way to the base of the race; even to the dogs, Max von Stephanitz, the founder of the GSD, used (click here). This will give the measurement of the inbreeding coefficient on “all generations”, as specified above, another dimension of course, but from genetic point of view, it is very interesting.

I don't know if this is because Aussies have a larger foundation group (300), but if pedigrees for CW start to fray at around 8-9 generations back for some dogs, I think that an 8-10 generation COI is probably pretty accurate

Partly true, depending on which Cvs you measure. If you measure the inbreeding coefficient of an older Cvs, then yes, as 8-10 generations in this case usually means that you measure all generations (to the base of the Breeding). If you do that for a young Cvs (newborn Cvs) it could mean that one simply disregard most part of the strong inbreeding effect of especially Rep z Pohranièní stráze.

Although we owe our beautiful breed in particular to Rep z Pohranièní stráze, it is unfortunate that large inbreeding with him was committed in an early stage and the initial gene pool quickly narrowed drastically. Therefore all our dogs are very much related to each other by him, with not really opportunities to outcrossing anymore. I don't know the initial breeding situation of the Australian shepherd, but apparently the basis of it consists from about 300 different dogs and the breed was probably not inbred right from such an early stage.This could explain the fact that the outcome of the inbreeding coefficient of 10 generations is not much different from that of more than 10 generations. This makes an important difference with our race, where extensive inbreeding in the beginning was committed. In regard to measurement accuracy, I therefore plead for measuring COI on all generations, at least to the base (> Rep z Pohranièní stráze included), as it was possible until recently. I have every trust is that soon it will be possible again.

Regards

[elf]

Another point, currently WD calculate the COI assuming the COI of common ancestors is 0 [Sum (1/2)^n1+n2+1], if you're using the complete formula introducing COI of common ancestors [Sum (1/2)^n1+n2+1 * (1+Fa)], you will see that calculating the COI with complete formula on 8 generations gives often a higher value than calculating the COI with troncated formula on full pedigree.

[Hanka]

Hi Buidelwolf, I have only one small note. Not all our dogs are related to Rep z Pohraniční stráže. We have 8 groups of males, but only 4 are based on him.
Yes, somewhere in mother side can be some Rep (or no) but in breeding we look at father lines.

[michaelundinaeichhorn]

Originally Posted by Hanka

Yes, somewhere in mother side can be some Rep (or no) but in breeding we look at father lines.

What I do not quite understand cause the genes of the mother are as strong as the genes of the father and obviously there.

Ina

[Hanka]

Of course. I thought, somebody will react. But in breeding of animals (horses, cows,....) we look at father lines more than on mother lines. But we must look at mother lines too.
But planning in breeding depends mor on father lines.

[elf]

This is interesting Hanka because this is the way breeding groups work, but it's indeed a bit disconnect with reallity. For exemple, let's take a dog from group V: Ali Terespol. Look at it's inbreeding composition on 8 generations: http://www.amicale-chien-loup-tcheco...d=6671&depth=8

As you can see Rep comes first with 8.04901123047%. Atos Heol CS comes 2nd with 1.77384614945%, so far less. It's the same for all groups.

In the other hand, calculating COI on full pedigree is very interesting but it's a statistical footprint, it's difficult to make the difference between the "old influence and the new one" at genetic level as each generation bring it's new contribution with epigenetic changes (among others) that we know transmissible.