Thanks for posting that last article--an excellent read.
Couple of questions about % blood--
As I understand it, values are assigned for each relationship, ie 25% for grandparent, 12.5% for great grand, then 6.25% and so on and so forth. Are the totals strictly cumulative? What if the ancestor in question is in both sire and dam's lines? Would it still be cumulative?
Also, typically how far back do you read a pedigree when planning a breeding? Back to the foundation dogs? In the example he gives in the article he takes it back 27 gens.
Lastly, I would imagine there is still a quotient of uncertainty in the results. With your years of experience coupled with pedigree analysis, how certain are you of your results with a new, unique breeding?
It is very complex, and there is an element of uncertainty. The tighter you breed, the higher the predictability. Homozygosity in the genetic coding means that there are more repeated genes which increases the likelihood of a trait being expressed (for good or ill). In Clumbers which are a breed with a smaller population, and a smaller gene pool, you will see more repetition of dogs, and therefore a higher degree of influence. In Goldens there are some very tightly linebred dogs still. When you go back into the history of the breed a bit there were some influential kennels where very intense linebreeding was practised, like Stenbury. They had a very recognizable outline. [The prefix is now used by someone else, who took over the breeding program after Ms. Minter's death in 1995--so look at pre-1995 dogs, and especially her dogs from the 40's to 60's to see how she built her line.]
It does accumulate when ancestors are repeated. So if a bitch is the grandmother and the great-grandmother there will be a higher contribution. COI does not tell all--influence tables go further into the degree of influence a particular dog has in the pedigree. So when I see a dog repeated over and over again, especially up to the fifth generation, that is a dog I want to investigate. That dog or bitch is having a greater influence because of their concentration in the genetic material. Remember though that each pup in a litter gets half its DNA from each parent--but not necessarily the same half. So that is why you can have variation in a litter, especially if it is more of an outcross. There are a couple of very influential dogs in show lines who are the results of breedings of litter sisters to litter brothers. Despite being double first cousins, they each set a very distinctive style when linebred on, and are associated with different health issues.
You can also get interesting outcomes in litters where there is some loose linebreeding. This is my litter from last spring Pedigree: Sterre Widgeon on The Wing
Each of the parents is the result of a breeding of an English/European pedigree sire, to a bitch from Trowsnest/Celestial lines. In common, they both brought Celestial Sirius Jake along. It has a fairly low COI, and I was actually doing the breeding to set something else up in the next generation, where there is a dog I like, but it did not make sense to take my current girls to him, but it does with Wings. Some of the pups take very much after the Jake part of the pedigree, some show elements of both, and others look super "English" and very much reflect the Camrose that is behind all of that part of the pedigree. The girl I kept, Wings, who is more on the Jake side of things, shows a gene expression that reflects the part of the pedigree I was breeding for in that combination (she is very like her paternal grandmother). Taking her to a male that represents that part of the pedigree down the road will more more likely (not a guarantee) produce similarity there, whereas it would be less likely with her sister who looks very English. The genes being expressed are at least in part an indication of what side of the parents pedigrees their particular gene combination came through. It is more complicated with recessive traits though, because those will not be expressed unless a gene is inherited from both parents--that is often where the interesting little surprises pop up--like a blue-eyed child in a family of brown-eyed siblings from two brown-eyed parents. Brown eyes are dominant in people, so the recessive blue eye gene could be hitching along for some time until the right combination is made and a child gets the blue gene from both parents.