Why pedigree and genomic COI can give different numbers
Two reports may give the same animal different coefficients of inbreeding (COI). That need not mean either report is wrong. One number may be calculated from recorded ancestors; another may be estimated from the animal’s DNA. Even two DNA-based numbers can differ if they use different methods.
This evidence review looks at three published studies of dogs and horses. It explains what their comparisons mean for a breeding decision. AnimalTrace did not collect the animals or data in these studies.
What each number measures
A pedigree COI estimates the chance that two copies of a gene in an animal came from the same ancestor, using the family tree on record. It depends on which ancestors are known and how far back the calculation goes. Two animals with the same parents receive the same pedigree expectation, though they do not inherit exactly the same stretches of DNA.
A genomic COI uses DNA markers or sequence data from the individual. One common approach measures runs of homozygosity: long stretches where the two copies of DNA match. Other approaches estimate inbreeding from marker frequencies or a genomic relationship matrix. The marker set and calculation matter, so “genomic COI” does not describe one interchangeable number.
For a basic introduction to the pedigree measure, start with our COI guide.
Pedigree depth changes the answer
In a Bearded Collie study, researchers calculated pedigree COI using all available generations, then only the most recent ten or five. Across the study’s pedigree cohort, the average fell from about 0.29 with the full pedigree to about 0.07 with five generations. Those are results for that population, not conversion factors for another breed. They show how many older shared ancestors a short calculation can miss.
A study across dog breeds likewise found that five-generation pedigree estimates did not capture the breed structure seen in deeper pedigrees and genomic measures. A low COI from a short family tree therefore needs its generation count beside it.
A good pedigree can still be useful
Genomic data add information about the individual animal, but a deep, well-maintained pedigree is valuable. Researchers studying the Pura Raza Española horse compared several pedigree and genomic methods in a population with unusually complete records. Their pedigree and genomic estimates were positively correlated, with correlations ranging from 0.58 to 0.79 for the comparisons they reported. The estimates still had different average values depending on the method used. The horse study shows why both the quality of the studbook and the choice of genomic calculation matter.
How to use two COI results
When the percentages disagree, ask four questions before treating one as the right number:
- How many pedigree generations were included, and how complete are they?
- What data and calculation did the DNA report use: a marker chip or genome sequence, and runs of homozygosity or another estimator?
- Were both numbers calculated for the same individual, rather than one being a breed average?
- Is the decision about a future pairing, the animal’s existing DNA, or the diversity of a whole breeding population?
A pedigree lets you inspect shared ancestors and compare possible matings before offspring exist. A genomic result can show the DNA the animal actually inherited. Neither percentage by itself identifies a disease variant or sets a universal “safe” threshold. Read it alongside the specific health tests and the goals of the breeding program.
Where AnimalTrace fits
AnimalTrace’s breeding checks show a pedigree COI based on the ancestors recorded in the product, with shared ancestors behind the number. The studies above do not validate an AnimalTrace calculation or establish a health outcome for any particular pairing.
Studies reviewed
This is a selective, plain-language review of three studies that directly compare pedigree and DNA-based measures. It is not a systematic review or original AnimalTrace research.
- Dreger and others (2016), Whole-genome sequence, SNP chips and pedigree structure in domestic dog breeds
- Perdomo-González and others (2022), Genomic and pedigree inbreeding rates in the Pura Raza Española horse
- Belanger and others (2025), Homozygosity and genetic diversity in the Bearded Collie