Beyond immediate family: relationships a parent-only check misses
Research note · Pedigree relationships
Abstract
Checking whether two animals have the same parent catches siblings. Checking whether one is the other’s parent catches another immediate relationship. Neither check follows ancestry farther back. We tested the reach of that shortcut in a published 237-dog research pedigree. Among the 207 sampled dogs with both parents recorded, there are 21,321 unordered pairs of animals. A full available-pedigree calculation found a nonzero recorded relationship for 10,386 pairs. Of those, 7,620 pairs (73.4%) had no direct parent–child link and no shared recorded parent. This is a property of a closely connected research family, not a breed-wide rate or a set of recommended matings.
Why look beyond a parent field?
A search that compares only parent IDs can quickly identify full or half siblings. It cannot find two animals connected through grandparents, earlier ancestors, or a longer route through the pedigree. That distinction matters when a breeder checks a prospective pair or a registry reviews a line: the absence of an immediate family link is not evidence that the recorded pedigree contains no relationship.
We asked a narrow, testable question: of the animal pairs with nonzero relationship in this source file, how many lack an immediate link? This extends our pedigree-depth study, which examined 30 historical sire–dam pairings, to every pairwise comparison among the 207 samples with two known parents.
Data and method
We used the sample IDs, sex codes, and parent IDs in the public pedigree file accompanying Campbell et al. (2016). All 237 sampled dogs remained in the ancestry graph. We selected the 207 with both a sire and dam recorded for pairwise comparison so the immediate-parent check would not mistake two missing parent fields for a shared parent. We compared each unordered pair once: 207 × 206 ÷ 2 = 21,321 comparisons. These include same-sex animals, ancestors and descendants, and historical animals that would never be a practical mating choice.
For each comparison, an immediate family link means that one animal is a recorded parent of the other, or that the two share a recorded sire or dam. We then calculated the tabular additive relationship across all ancestry available in the file. A value above zero means the recorded graph contains a relationship; zero means no relationship was traced within that graph. Unrecorded founders are treated as unrelated for this calculation. We used no SNP genotypes, disease results, or dates.
Results
| Pairwise category | All comparisons | Nonzero recorded relationship | Zero within recorded pedigree |
|---|---|---|---|
| Direct parent–child or shared parent | 2,766 | 2,766 | 0 |
| No immediate family link | 18,555 | 7,620 | 10,935 |
| All pairs of the 207 sampled dogs | 21,321 | 10,386 | 10,935 |
Thus 7,620 of the 10,386 pairs with nonzero recorded relationship—73.4%—would not be identified by the immediate-link rule alone. Looked at from the other direction, 41.1% of pairs with no immediate link still have a nonzero relationship in the available pedigree. These are pairwise comparisons, not independent observations: one dog appears in many rows.
One source pairing makes the difference concrete. PFZ23A05 and PFZ23A07 have no shared recorded parent and are not parent and child. Both trace to PFZ26F05 through a parent, making that dog a shared recorded grandparent. Their calculated additive relationship is 6.25%; the expected offspring pedigree COI for this historical sire–dam pairing is 3.125%. A check limited to the two animals’ immediate parents would not find that shared grandparent.
What the finding means for AnimalTrace
Search is useful for finding and verifying the two animal records. For a pairing question, Match needs the connected pedigree, not only the two parent fields. A result is more useful when it can identify the recorded ancestor behind a nonzero value and reveal where ancestry stops. Our worked AnimalTrace example shows Match tracing a different pairing in this same public family to PFZ26F05.
The same logic applies after a parent record changes. A direct parent check finds the edited link; deeper ancestry determines which other pairwise calculations may need review. Our single-link sensitivity study measured that effect on the 30 recorded sire–dam pairings. The present analysis measures the broader relationship graph, regardless of whether two animals were ever mated.
Limits and reproducibility
This research colony was selected for genetic study and has a particular family structure. The 73.4% figure must not be generalized to a breed, registry, or proposed mating pool. The 21,321 comparisons include relatives, same-sex pairs, and cross-generation pairs; they are not 21,321 breeding candidates. Missing ancestry can hide real biological relationships behind a calculated zero. A nonzero pedigree relationship is not a health diagnosis, a genomic relationship measurement, or a recommendation about a mating. We report counts rather than a significance test because the comparisons reuse the same animals and were not drawn as an independent population sample.
- Source study: Campbell et al., 2016, A Pedigree-Based Map of Recombination in the Domestic Dog Genome.
- Source data: Authors’ public repository,
new_pedigree_dogs.faminsidedog_genotype_data.tar.gz. - Derived data: Three-row category summary and all 21,321 pairwise comparisons.
- Method: Download the standard-library Python script to repeat the comparison and relationship calculation.
This AnimalTrace research note has not undergone external peer review.