COI (Inbreeding Coefficient based on Wrights formula)

Coefficients of Inbreeding, or COIs for short, are calculations made on a pedigree to determine a rough estimate of how inbred an individual may be. COIs can be calculated a number of ways. These different formulas are generally named after the statistician or geneticist who derived them (Wright, Hardiman, Meuwissen, etc.) The difference between most of these formulas has to do with how well-defined the population is on which the calculation is being performed.
Wright's Inbreeding Coefficient is the one most used by purebred dog breeders, but there are others.

Inbreeding coefficients are generally calculated using the formula:Â

FX =½RSD

where F is the calculated coefficient of inbreeding, X is the individual of interest, and RSD is the relationship between the sire and dam.

In order for a litter to be inbred, both sire and dam have to have one or more dogs in common between their pedigrees. Even though one or the other of the parents may be very inbred themselves, if the dog they are mated to does not have any ancestors in common with the inbred parent, the resulting puppies will not be inbred, and the COI will be zero. This rarely happens in purebred dog populations, as most are founded on a few animals whose names are repeated sometimes far back in a pedigree.

When a sire and dam do have some ancestors in common, the amount of inbreeding of those ancestors has must be included in calculating the COI. For instance, if Dog X appears twice behind the sire, and twice behind the dam, it will affect the COI. If Dog X is not inbred, this will result in a lower COI than if Dog X is inbred himself. The more inbred an ancestor is, and the more often that ancestor appears behind both the sire and the dam, the higher the resulting COI will be. To accurately calculate COIs in purebred dogs, the COI of any ancestor must be known, and factored into the calculation. This means that not only must you calculate how many times each dog appears on both sides of a pedigree, you must also calculate the COI for each common ancestor. This is shown by the calculation:

FX = ∑ [(1/2)P1 + P2 +1 (1 + FA)]

Where P1 is the number of generations from one parent back to the common ancestor, and P2 is the number of generations from the other parent back to that same ancestor. FA is the COI for the common ancestor.

As you can see, knowing the COI of the ancestors is very important, particularly in breeds which have had bottle necking episodes in their pasts, or have never been very numerous. This also applies especially today, given the popularity of line-breeding in all breeds. Line-breeding is another form of inbreeding, and results in cumulative increases in COIs over the generations.

It is not enough simply to look at the ancestors in a 3, 4, or 5-generation pedigree. Those dogs appearing in a 5-generation pedigree may have many ancestors in common. This needs to be taken into consideration if one is using COIs as a breeding tool. Calculating COIs to the tenth generation (at least) is recommended by many population geneticists familiar with dog breeding. Taking this much wider slice of the population when calculating COIs allows the breeder to have a better picture of the true inbreeding coefficient behind any dog or planned mating.

One important thing to keep in mind when reviewing any dog's COI, is that the COI is calculated based on what appears on paper. It does not represent actual analysis of a dog's DNA. Since COIs are calculated solely by looking at names on a pedigree, they are probabilities only, not absolutes. The COI predicts the likelihood that any one allele may have come from a particular ancestor, or the likelihood that an individual's genetic makeup comes from one or more particular ancestors, thus increasing the likelihood that the individual may be homologous for one or more traits. A COI is not a guarantee of future greatness; neither is it a warning of potential deleterious effects of inbreeding.

No COI stands alone. A COI is only valuable when it can be compared to the population as a whole. One cannot tell if a COI is "high" or "low" unless one knows what is the norm, mean, median, or average COI for a particular breed. Additionally, only COIs calculated in an identical manner should be compared. Click here to see examples of how COIs calculated on the same individual, but using different methods, can result in very different values.

Today, there are many pedigree-generating software packages available, and many COI calculators available for dog breeders to use. Many pedigree websites (including this one), and even some official registries are now displaying COI percentages at the dogs page in their databases. Some registries and breed societies have recommendations and restrictions regarding how high COIs are permitted to be for individuals, and for planned matings. One should exercise caution when comparing COIs, as each website's calculator may be calculating COIs using a different formula, and each site or registry may be calculating to a different number of generations. It is always best to ask the webmaster or breed warden of a particular site to send you the brief formula (such as the ones on this page) used for the calculations on the site, what each symbol in the formula means, and how many generations their COIs are calculated to. It is also good to ask if the COIs of common ancestors are used in their site's calculations. Some COI generators do not include ancestral COIs in their calculations, which can result in significantly lower COIs than actual. This information is especially important to have from those registries which have placed restrictions on breeders using their registry services or importing dogs. Because of this wide variation in COI accuracy, this site recommends that COIs not be used as the sole, or even a major, determining factor when deciding whether to do a particular mating.

Article Source: chessieinfo.net

A huge thanks to the persons mentioned below. You have all contributed to the COI script being available at SBTpedigree.com.

COI donators:

  • Silje Stakston (Denmark)
  • Sonja Porits (Austria)
  • Julia Ehmann (Austria)
  • Mette Grøntoft (Norway)
  • Felipe Cobério Moreira (Brasil)
  • Pia Helen Olsen (Norway)
  • Katrine Weserby (Norway)
  • Anita Steen (Norway)
  • Ken Cox (Australia)
  • Edita Omulec Ivas ()
  • Andrea Bonato (Italy?)
  • Thais Castro (Brazil)
  • Tadeu Alves (Brazil)
  • Cristina Cozza (Italy))
  • Michael Smith (Australia ?)
  • Jennifer Rice (USA)
  • T M Walker (UK)
  • Mitja Grah (Slovenia?)
  • Panos Maritsas (Denmark)
  • Tomislav Matulic (Croatia)
  • Samira Gzom (Belgium)
  • Roberta Caruso (Italy)
  • Gian Luca Meles (Italy)

On behalf of all the SBTpedigree.com users - Thank you for your contribution.