Anyone who looks into breeding a litter soon meets the abbreviation COI. The coefficient of inbreeding is a single number meant to summarize how closely related a puppy's parents are, and by extension how much of the puppy's DNA is likely to be identical on both sides. It is a useful number, but it is easy to misread. This guide explains what it measures, how pedigree and DNA-based figures differ, and why it matters even for dogs that are clear on every single-gene test.
Key facts
- COI is the probability that the two copies of a gene in a dog are identical because they come from the same ancestor.
- A full-sibling or parent-offspring mating gives a pedigree COI of 25%; half siblings 12.5%; first cousins 6.25% (assuming unrelated ancestors beyond that).
- Genomic COI, measured from DNA, is usually much higher than a pedigree COI based on a few generations, because it captures relatedness from far back.
- Across 227 breeds, the average genomic inbreeding in one large study was 0.249, and more inbred breeds had more veterinary care events (Bannasch et al., 2021).
What COI actually measures
Every dog has two copies of each gene, one from the sire and one from the dam. If the parents share ancestors, there is a chance that a puppy inherits the same ancestral copy through both parents. The coefficient of inbreeding, introduced by the geneticist Sewall Wright in the 1920s, is the probability that this happens at any given place in the genome. A COI of 25% means that, on average, a quarter of the puppy's genome is expected to be "identical by descent".
That matters because many harmful variants are recessive and only cause disease when a dog has two copies. The more of the genome is identical by descent, the more likely a puppy is to receive two copies of a rare harmful variant that sits hidden in carrier ancestors. Inbreeding also tends to reduce fertility, litter size and immune diversity, an effect known as inbreeding depression.
Typical values for common matings
For a pedigree COI, assuming the ancestors further back are unrelated:
| Mating | COI of the puppies |
|---|---|
| Parent × offspring, or full siblings | 25% |
| Half siblings, grandparent × grandchild | 12.5% |
| Uncle × niece, aunt × nephew | 12.5% |
| First cousins | 6.25% |
| Second cousins | about 1.6% |
In real pedigrees the ancestors are rarely unrelated, so repeated use of the same lines pushes the figure higher than the simple table suggests.
Pedigree COI versus genomic COI
Pedigree COI is calculated from the family tree. Its value depends heavily on how many generations are included. A five-generation COI may look low simply because the common ancestors that made the breed sit eight or fifteen generations back. That is why kennel clubs that publish COIs, such as the Kennel Club in the UK through its Mate Select service, also publish the breed average, so a mating can be compared with the breed as a whole rather than with an absolute target.
Genomic COI is measured directly from DNA, usually from a genotyping array that reads tens or hundreds of thousands of markers. It estimates how much of the dog's genome actually is identical, including long stretches of homozygosity inherited from the breed's founders. Genomic figures are typically much higher than pedigree figures. In the study by Bannasch and colleagues, which used DNA data from 49,378 dogs of 227 breeds, the average genomic inbreeding was 0.249, the level expected from a full-sibling mating in an unrelated population.
The two numbers answer different questions. Pedigree COI is good for comparing possible matings within a breed. Genomic COI shows what the individual dog really inherited, since two full siblings can have quite different genomic values by chance.
Does inbreeding affect health?
The same 2021 study linked insurance-based morbidity data to the inbreeding figures. Breeds with low and high inbreeding differed significantly in morbidity, and in a regression model both body size and inbreeding were significant: smaller, less inbred breeds tended to be healthier than larger, more inbred ones. Brachycephalic breeds were analyzed separately because their shape-related problems would have distorted the results. Several of the authors worked for a company that sells dog DNA tests, which the paper declares; the findings agree with a long body of research on inbreeding depression in animals.
Inbreeding is a breed-level problem as much as a litter-level one. When a few popular sires produce a large share of a breed's puppies, the effective population size falls and the whole breed becomes more inbred within a few generations. Genoscoper's article on popular sire syndrome and genetic diversity and its Kromfohrländer diversity study show how DNA data was used to measure this in real breeds.
How breeders use COI
- Compare with the breed average. A common guideline is to plan matings at or below the breed's average COI, which over time slows the rise in inbreeding.
- Combine with DNA tests. Single-gene tests (such as those for PRA or MDR1) tell you about known variants. COI tells you about the unknown ones. A mating can be clear on every test and still carry high inbreeding risk.
- Use more males. Spreading litters across many sires keeps the effective population size up; many breed health strategies limit the number of litters per sire.
- Do not discard good carriers. Excluding every carrier of a common recessive variant can shrink the gene pool and raise inbreeding. Breeding a carrier to a tested clear partner avoids affected puppies while keeping the dog's other qualities in the breed.
In Finland, breed clubs set their own breeding goals in a breed-specific breeding strategy (JTO) approved by the Finnish Kennel Club, and the Kennel Club's KoiraNet breeding database shows pedigree inbreeding figures for registered dogs.
Frequently asked questions
What is a good COI for a dog?
There is no single safe number, because breed averages vary widely and pedigree depth changes the figure. A practical rule is to aim below the average for the breed, using the same number of generations for every comparison.
Why is my dog's DNA-based COI so much higher than its pedigree COI?
Genomic COI captures relatedness from all past generations, including the breed's founders, while a pedigree COI usually covers only five to ten generations. Both can be correct; they measure different depths.
Do mixed-breed dogs have a COI?
Yes. A genomic COI can be measured for any dog. First-generation crosses between unrelated breeds usually have very low values.
Is a low COI a guarantee of health?
No. A low COI lowers the risk of recessive disease and inbreeding depression in general, but a dog can still inherit a specific disease variant from carrier parents. That is what DNA tests are for.
Sources
- Bannasch D, et al. (2021). The effect of inbreeding, body size and morphology on health in dog breeds. Canine Medicine and Genetics 8:12. doi:10.1186/s40575-021-00111-4.
- Wright S (1922). Coefficients of inbreeding and relationship. The American Naturalist 56:330–338.
- The Kennel Club (UK): Mate Select and breed COI averages.
- Suomen Kennelliitto (Finnish Kennel Club): KoiraNet breeding database and breed-specific breeding strategies (JTO).