Coat Colour Genetics for Dog Owners

A dog's coat colour and pattern are decided mainly by a set of gene loci, each with its own variants: A, B, C, D, E, K, S and M, plus modifiers for intensity, harlequin, white markings, ticking, brindle, sable, merle and parti-colour. These loci do not act alone; epistasis means one locus can mask or override the expression of another. Prediction tools and DNA panels answer different questions, and the pairings an owner chooses have consequences that go past appearance.

Which loci decide a dog's coat colour and pattern?

Coat colour in dogs is not one gene but a series of loci, each named with a letter, and each carrying variants that interact. The A locus governs the distribution of black and red pigment in the coat, producing patterns such as sable, agouti and tan points. The B locus controls black versus brown pigment, while the D locus controls dilution, turning black into blue and brown into lilac. The E locus decides whether a dog can produce black pigment in the coat at all, and the K locus sits above it, determining whether brindle or solid black is expressed. The S locus and its related white-marking genes control how much white appears, from a small chest patch to an almost entirely white dog. The M locus produces merle, a pattern of mottled patches that varies widely in extent.

Beyond these, modifier genes fine-tune the result. Intensity genes adjust how deep a red or yellow coat appears, harlequin modifies merle in some breeds, and ticking adds small coloured spots to white areas. Because these loci interact through epistasis, the visible colour of a dog is the sum of several decisions, not the output of a single switch. A dog can carry a variant for a colour it does not show, and that hidden variant can reappear in the next generation. Breeders and students who want the full locus-by-locus breakdown, including how each one masks the next, can find it at canine coat colour genetics, which covers the A, B, C, D, E, K, S and M loci and the modifiers between them.

How do prediction tools and DNA panels differ?

Prediction tools and DNA panels are often treated as competing methods, but they answer different questions and work best together.

A Punnett square is the classic prediction tool. It takes the known or assumed genotypes of two parents and lays out the possible combinations for their offspring, giving a probability for each outcome per litter. The same logic underlies probability calculations by litter size: a pairing that gives a one-in-four chance of a colour will not produce that colour in every litter of four, and the tool shows the range rather than a guarantee. Punnett squares depend entirely on the genotypes entered. If the assumed genotype of a parent is wrong, the prediction is wrong with it.

DNA panels test the dog itself. A swab or blood sample is analysed for specific variants at named loci, and the result reports which variants the dog carries. Panels differ in which loci they cover, how many variants they test at each locus, and how they report uncertain results. Comparing laboratories is part of the work: two panels may test the same locus but include different variants, so a clear result from one is not automatically equivalent to a clear result from another. A panel can confirm a carrier state that a Punnett square could only estimate, and a Punnett square can show what a panel result means for a planned litter. Neither replaces the other.

What are the breeding consequences of colour pairings?

Colour pairings carry consequences that are not visible in the coat. Two are well documented and widely discussed among breeders.

The first concerns dilution. When two dogs that both carry dilution are paired, the resulting puppies can be double dilute, and double dilute dogs are associated with colour dilution alopecia, a condition in which the hair becomes sparse or falls out. The coat colour itself may be within the breed standard, but the health of the coat is not guaranteed by the colour.

The second concerns merle. Breeding two merle dogs together can produce double merle puppies, which are often largely white and carry a higher risk of deafness and eye abnormalities. The same logic applies to double dilute pairings in breeds where dilution is common. These outcomes are not rare accidents; they follow directly from the genotypes of the parents, which is why a panel result matters before a pairing is planned rather than after.

Colour and health correlations also vary by breed. Some breeds have specific colour-linked conditions that appear only in certain patterns, and a pairing that is safe in one breed may carry different risks in another. Breed-specific guidance is therefore part of any breeding plan, not an optional extra.

Why hidden variants matter in a breeding plan

A dog that shows a solid colour can still carry variants for patterns it does not express. Because of epistasis, a recessive variant at one locus can be masked by a dominant variant at another, and the dog appears to be something it is not, genetically. This is the reason two apparently similar dogs can produce very different litters.

For a breeder, the practical consequence is that visible colour is a weak guide to genotype. A dog that has never produced a particular colour may still carry it, and a dog that has produced it once will continue to carry it. Testing removes the guesswork, and a Punnett square then turns the test result into a probability for the planned litter. The two steps belong together: test the parents, then calculate the outcomes.

How should a breeder use colour information?

Colour information is most useful when it is treated as one input among several, not as the goal of a breeding programme.

A workable sequence is to establish the genotype of each potential parent with a panel that covers the loci relevant to the breed, to compare laboratories so the results are read on the same terms, and to run the probabilities for a proposed pairing before committing to it. Where a pairing raises the risk of double merle or double dilute, the pairing can be changed, or the risk can be accepted with full knowledge of what it means for the puppies.

Colour genetics is also a teaching subject in its own right. The loci, the epistatic relationships and the probability calculations form a coherent system that can be learned from first principles, and breeders who understand the system make fewer assumptions about what a dog carries. The same body of material serves owners who simply want to know why a litter turned out the way it did.

What a colour result does not tell you

A colour panel result describes the variants tested at the loci covered by that panel. It does not describe the dog's overall health, temperament or suitability as a breeding animal, and it does not cover loci the panel omits. A clear result at one locus says nothing about another.

Breeders who treat a panel as a complete picture of a dog are reading more into the result than it contains. The useful reading is narrower and more precise: these are the variants this dog carries at these loci, and here is what those variants mean for a proposed pairing. That is enough to plan a litter with fewer surprises, and it is the level of certainty the tools actually provide.

Border Collie coats are described through a set of visible pigments and patterns, including black, brown, blue and lilac dilutes, merle, brindle, sable, tan points, red ee and white. Markings add detail to that description: a blaze is a white stripe running down the face, ticking refers to small flecks of colour scattered through white areas, and other terms cover a white collar or roan shading. For readers who want the full vocabulary, a practical coat colour guide lists the colours and marking terms in plain language. One rule from that entry matters here: keep the written description and the genetic genotype as two separate records, since what the eye sees does not confirm which genes a dog carries.