Any Ideas on Interpreting Inbreeding Coefficient (IC) Results? - Page 1

i found a free online Inbreeding Calculator http://www.czerwonytrop.com/inb/index.php?full=ok&lng=en
so just for fun i input a 7 generation pedigree just to see what would come up, but i'm not clear on how to read the results. this is the top portion of the results; does anyone have any ideas and/or thoughts on these results and the dogs listed?- tia

Coefficient Results

Inbreeding: 0.0296630859375 (2.96630859375%) AVK (ancestor loss coefficient): 76.3779527559% The computation has been done for 7 generations

Inbreeding contribution by acenstor
BERRY VOM NEUEN LANDE	0.015625 (1.5625%)
GILDO VOM KöRBELBACH	0.00634765625 (0.634765625%)
FADO VON KARTHAGO	0.00390625 (0.390625%)
GREIF ZUM LAHNTAL	0.0029296875 (0.29296875%)
AFRA VOM STOPPENBERGER LAND	0.00048828125 (0.048828125%)
CARO VOM ALLERSWALD	0.0003662109375 (0.03662109375%)

i thanks for taking the time to explain what COI is.  just out of curiosity about this topic i had been looking for a program that calculated COI but they are all for windows so for those of us who don't have a windows pc this site provides a nice alternative.

i did read the site so i do understand the %'s themselves; and from what it says it would seem this mating would be pretty nice from a COI pov.  and if i understand this whole thing correctly this method does not factor in the inbreeding done prior to the 7 generation calculated, and also excludes the inbreeding done in the breeds foundation dogs. i'm not sure one can ever know the true COI but even with these limitations i 'm wondering if this info might still prove to be useful; i'm not quite sure how. 

what i'm most curious about is how this sort of info could be used, not only in helping choose a mating that  could create more or less homozygosity, but i'm wondering if this sort of info could prove to be a useful tool in other ways. perhaps it could provide some insight into the types of pups a particular mating might produce?

so i guess my biggest question is:  is it possible to learn more about the potential of a litter by knowing the COI, particularly something about the dogs listed as Inbreeding contributors? in the example i originally posted, these dogs popped up:
- Berry vom neuen Lande
- Gildo vom Körbelbach
- Fado von Karthago
- Greif zum Lahntal
- Afra vom Stoppenberger Land
- Caro vom Allerswald
could knowing the physical & mental abilities, and temperament of these dogs help in any way with understanding the pups/dogs they contributed to? and could knowing the percentage of their contribution be particularly helpful? this is all just out of general curiosity - tia

If you are interested in a general way in understanding genetic diversity in a population and linebreeding effects, take a look at this article by Fred Lanting. http://www.siriusdog.com/inbreeding-linebreeding-diversity-genes.htm

You will probably have to cut and paste since my browser isn't completely compatible with the editor here and doesn't put in the clickable hyperlinks.

COI estimates are just that, estimates. But as with any quantitative study, the results may be fairly close to the COI estimates, because the probability of their homogenous state (and therefore expression) will tend to gravitate toward the peak of the bell curve (the sum of its overall homogenous allele pairs). This can eventually result in the expression of deleterious genes that were before unknown in the considered common ancestor, not being expressed in the phenotype of either the ancestor or its direct progeny. But also can result in a captured genotype that is more consistent of producing the phenotype of that ancestor.

Going back only 7 generations is practical, but a limited view considering the overall breed's standing COI. There have been many significant bottlenecks, the first beginning with Horand, and some recent ones that draw much concern, though shouldn't necessarily. Fero, for example, though he's certainly not the first in the breed's history. What particular lines through Fero ever proved faulty for linebreeding? If any, perhaps those particular progeny should be avoided, and not so much necessarilly for any reason related to Fero as their damside. You're talking about a genetic snapshot in time that has proven over and over again, that linebreeding on the "good" ain't all that bad. There are other "greats" in the breed, that I wouldn't personally support that statement with, of course. But the point being, if there's no harm, and in fact proves consistently well when linebred on, then there's no substantial reason to avoid such a dog, even in both of your breeding pairs.

As the COI estimates homogenous pairs of alleles in the dna structure, there is also a budding genomic science that emphasizes actual readings of SNP's, or heterogenous pairs of alleles. This science measures the actual genotypes of the decoded genome structures, rather than statistical estimates. You can read more and download free SNP genotyping software at www.appliedbiosystems.com Currently, their readymade software only includes human or mouse genomes, but state that others are available upon request. The canine genome was decoded a few years ago, but I doubt you could find breed specific information at this point.

Eventually, I expect you will be able to compare sampled dna data of your actual breeding stock to the overall breed, and calculate heritability estimates of their combined breeding on your personal computer. But until a better understanding is grasped of this potential by the general public (esp. gsd fanciers), then there's really no point. The common irrational hysteria about inbreeding's detriments and ALL the supposed benefits of genetic "diversity" is going to ruin the breed before it's worth saving. This mentality is like saying "forks are bad, only spoons are good", when in fact, they simply serve a different purpose, unrelated of any ethical standpoint.

I guess this old windows 98 system can't speak paragraphs [fixed*]

The common irrational hysteria about inbreeding's detriments and ALL the supposed benefits of genetic "diversity" is going to ruin the breed before it's worth saving. This mentality is like saying "forks are bad, only spoons are good", when in fact, they simply serve a different purpose, unrelated of any ethical standpoint.

thanks this is interesting. i'm an arrant dilettante here but i get what you are saying and completely agree that people react to the idea of inbreeding without truly understanding what are the exactly pros and cons. inbreeding is not the real problem; it can lead to problems or reveal problems, but is not the actual problem.  however it is not something that should be flippantly practiced either. imo the problems the gsd breed is experiencing now is only partially due to concentrating on a very few dogs; another huge issue is breeding for what's in fashion at the expense of true quality - temp, workability & intelligence - we are lossing very good genes for the sake of trends.  imo it's important to realized that there have been and can be some very close breedings which produce wonderful results, while others are utterly disastrous.  i think  perhaps a good rule of thumb is: if you do not know the pedigrees of the dogs you are breeding inside and out you are flat out gambling, so to hedge your bet stick to out-crossing and avoid linebreeding. this way you aren't concentrating any recessive genes you are not aware your dogs carry. 

maybe this focus/pendulum swing from the trend of inbreeding of 50 yrs ago to the great push now to out-cross, to keep the inbreeding % low, is to accommodate a culture of "breeding for dummies".  but this aside if we assume that the breeder is serious and wants to master their chosen profession, wants to learn their dogs pedigrees inside and out - would knowing the COI of just 7 generations and looking at and learning about the dogs that have been concentrated (the inbreeding contributors) help with designing a training program and or handling style that could/would maximize the dogs potential?  would knowing the ratio between the dogs serve any real purpose? or is the info gleaned from COI simply too general or theoretical to provide much practical insight?  again i'm just trying to understand if there is much direct practical application for the information gotten from this way of compiling pedigree data.

The problem of COI's, is that they are an examination of the probability of homologous allele pairing of an overall entire genome from a single dog, and not the estimation of a specific allele pairing from an account of all genetic contributors of the pedigree. If you knew specifically, what the said ancestor carried, and you wished to either avoid or assimilate that into your breeding, not having information of whether or not the direct progeny of that ancestor carries the gene, then it could help, but still not provide a full account.

I have a dog, linebred 3-3,4 on Yoschy (a mere 15-5/8ths% COI). Yoschy is known to carry the longcoat gene, but my risk increases if I can verify that a particular son or daughter carries the same, and decreases if I can verify that his direct progeny undoubtedly do not carry the gene. But then, I must ask, how many other's in the pedigree carry that recessive, and how close up? COI's can quickly become irrelevant, if I have no way to account for Yoschy offspring or non-related dogs who carry it.

Since COI calculations are quantitative, and many/most characteristics are polygenic, this could be more relevant in a closer linebreeding, such as a 3-2 on a line of strong hip producers. Fewer generations passed, translates to greater retention of the "group set", and less noise from other contributers.

If there's gaps in the knowledge of the pedigree, then any information could help in your estimation I suppose, but the most practical and effective way to produce is first and foremost by selection of clearly evident phenotype, from contributions shared by their direct predecessors. (This inevitably will lead to popular producers, and therefore linebreeding) Then a breeder can concern themselves with the mysteries of latent genes, or prepotency for type. If one begins with dissimilar and/or less than average breeding stock, every positive characteristic reinforced will be countered with an even larger share of the worst.

Until a model illustrating the complex interactions of temperament and it's heritability becomes available, even the future of SNP genotyping is of limited use, in my opinion. Yes, we can see the genes, but what do they mean, specifically? Some genes even seem to provide tradeoffs, alternate types being of significant benefit and each single type being multifaceted, given their circumstances, as in the example here: www.ehretgsd.com/COMT.pdf