Horse Coat Color Inheritance Planner

See how parental genotype assumptions affect modeled horse coat colors. This planner combines red-factor and agouti interactions with cream dilution and gray-variant inheritance. For uncertain parental genotypes, it evaluates every permitted pairing and shows the minimum and maximum probability for each outcome, without inventing a carrier frequency or treating all unknown genotypes as equally likely. Inspect a specific pairing, trace a gamete combination, select a union of target outcomes, and export exact probabilities with their assumptions. It covers four allele classes rather than every horse color gene or real-world appearance.

Coat inheritance lab

See the possibilities behind a coat color

Trace red factor, agouti, cream and gray inheritance. Keep exact probabilities for known pairings separate from ranges caused by unknown parental genotypes.

Start with explicit genotype assumptions

This is a four-locus educational model, not a complete coat prediction or a breeding suitability assessment. Other dilution and spotting genes can change appearance. Gray means a modeled variant is inherited; it does not mean a foal is born gray.

Choose the outcomes you want to understand

Select any combination as one target event. A foal matching any selected class counts once. “Gray variant present” retains the underlying base/dilution class; actual appearance can change with age.

Selected pairing target probability

28.125%

Exact fraction 9/32

Across 1 possible pairings

28.125% – 28.125%

Range of possible probabilities, not a confidence interval or an average.

Inspect one possible parental pairing

This pairing is an illustrative member of your candidate set, not a most-likely pairing. Candidate genotypes have no assigned probabilities.

Red factor E/r · Agouti A/a · Cream N/Cr · Gray variant class N/N

Red factor E/r · Agouti A/a · Cream N/Cr · Gray variant class N/N

This pairing: probability distribution

These 18 exclusive outcomes sum to 100%. Each bar uses a fixed 0–100% scale.

chestnut · no modeled gray variant6.25% · 1/16
chestnut · gray variant present0% · 0
palomino · no modeled gray variant12.5% · 1/8
palomino · gray variant present0% · 0
cremello · no modeled gray variant6.25% · 1/16
cremello · gray variant present0% · 0
bay · no modeled gray variant14.0625% · 9/64
bay · gray variant present0% · 0
buckskin · no modeled gray variant28.125% · 9/32
buckskin · gray variant present0% · 0
perlino · no modeled gray variant14.0625% · 9/64
perlino · gray variant present0% · 0
black · no modeled gray variant4.6875% · 3/64
black · gray variant present0% · 0
smoky black · no modeled gray variant9.375% · 3/32
smoky black · gray variant present0% · 0
smoky cream · no modeled gray variant4.6875% · 3/64
smoky cream · gray variant present0% · 0

All candidate pairings: probability bounds

Each interval runs from its own minimum to maximum on a 0–100% scale. Endpoints may come from different pairings; do not add them as a distribution.

chestnut · no modeled gray variant: 6.25% – 6.25%

chestnut · no modeled gray variant: 6.25% to 6.25% across candidate pairings.

chestnut · gray variant present: 0% – 0%

chestnut · gray variant present: 0% to 0% across candidate pairings.

palomino · no modeled gray variant: 12.5% – 12.5%

palomino · no modeled gray variant: 12.5% to 12.5% across candidate pairings.

palomino · gray variant present: 0% – 0%

palomino · gray variant present: 0% to 0% across candidate pairings.

cremello · no modeled gray variant: 6.25% – 6.25%

cremello · no modeled gray variant: 6.25% to 6.25% across candidate pairings.

cremello · gray variant present: 0% – 0%

cremello · gray variant present: 0% to 0% across candidate pairings.

bay · no modeled gray variant: 14.0625% – 14.0625%

bay · no modeled gray variant: 14.0625% to 14.0625% across candidate pairings.

bay · gray variant present: 0% – 0%

bay · gray variant present: 0% to 0% across candidate pairings.

buckskin · no modeled gray variant: 28.125% – 28.125%

buckskin · no modeled gray variant: 28.125% to 28.125% across candidate pairings.

buckskin · gray variant present: 0% – 0%

buckskin · gray variant present: 0% to 0% across candidate pairings.

perlino · no modeled gray variant: 14.0625% – 14.0625%

perlino · no modeled gray variant: 14.0625% to 14.0625% across candidate pairings.

perlino · gray variant present: 0% – 0%

perlino · gray variant present: 0% to 0% across candidate pairings.

black · no modeled gray variant: 4.6875% – 4.6875%

black · no modeled gray variant: 4.6875% to 4.6875% across candidate pairings.

black · gray variant present: 0% – 0%

black · gray variant present: 0% to 0% across candidate pairings.

smoky black · no modeled gray variant: 9.375% – 9.375%

smoky black · no modeled gray variant: 9.375% to 9.375% across candidate pairings.

smoky black · gray variant present: 0% – 0%

smoky black · gray variant present: 0% to 0% across candidate pairings.

smoky cream · no modeled gray variant: 4.6875% – 4.6875%

smoky cream · no modeled gray variant: 4.6875% to 4.6875% across candidate pairings.

smoky cream · gray variant present: 0% – 0%

smoky cream · gray variant present: 0% to 0% across candidate pairings.

Trace a gamete pairing

Alleles are ordered red factor / agouti / cream / gray. Identical gametes are combined. Cell probabilities multiply their two header probabilities; equally sized cells are not necessarily equally likely. A check mark identifies a selected target outcome.

Parent 1 gametes × parent 2 gametes
Parent 1 ↓ / Parent 2 →r / a / N / N
1/8
r / a / Cr / N
1/8
r / A / N / N
1/8
r / A / Cr / N
1/8
E / a / N / N
1/8
E / a / Cr / N
1/8
E / A / N / N
1/8
E / A / Cr / N
1/8
r / a / N / N
1/8
r / a / Cr / N
1/8
r / A / N / N
1/8
r / A / Cr / N
1/8
E / a / N / N
1/8
E / a / Cr / N
1/8
E / A / N / N
1/8
E / A / Cr / N
1/8

Selected cell: row 1, column 1

chestnut · no modeled gray variant

Red factor r/r · Agouti a/a · Cream N/N · Gray variant class N/N.

1/8 × 1/8 = 1/64 = 1.5625%. Outside the selected target.

Offspring genotype ledger for this pairing
GenotypeModeled coat classExact probability
Red factor r/r · Agouti a/a · Cream N/N · Gray variant class N/Nchestnut · no modeled gray variant1/64 · 1.5625%
Red factor r/r · Agouti a/a · Cream N/Cr · Gray variant class N/Npalomino · no modeled gray variant1/32 · 3.125%
Red factor r/r · Agouti a/a · Cream Cr/Cr · Gray variant class N/Ncremello · no modeled gray variant1/64 · 1.5625%
Red factor r/r · Agouti A/a · Cream N/N · Gray variant class N/Nchestnut · no modeled gray variant1/32 · 3.125%
Red factor r/r · Agouti A/a · Cream N/Cr · Gray variant class N/Npalomino · no modeled gray variant1/16 · 6.25%
Red factor r/r · Agouti A/a · Cream Cr/Cr · Gray variant class N/Ncremello · no modeled gray variant1/32 · 3.125%
Red factor r/r · Agouti A/A · Cream N/N · Gray variant class N/Nchestnut · no modeled gray variant1/64 · 1.5625%
Red factor r/r · Agouti A/A · Cream N/Cr · Gray variant class N/Npalomino · no modeled gray variant1/32 · 3.125%
Red factor r/r · Agouti A/A · Cream Cr/Cr · Gray variant class N/Ncremello · no modeled gray variant1/64 · 1.5625%
Red factor E/r · Agouti a/a · Cream N/N · Gray variant class N/Nblack · no modeled gray variant1/32 · 3.125%
Red factor E/r · Agouti a/a · Cream N/Cr · Gray variant class N/Nsmoky black · no modeled gray variant1/16 · 6.25%
Red factor E/r · Agouti a/a · Cream Cr/Cr · Gray variant class N/Nsmoky cream · no modeled gray variant1/32 · 3.125%
Red factor E/r · Agouti A/a · Cream N/N · Gray variant class N/Nbay · no modeled gray variant1/16 · 6.25%
Red factor E/r · Agouti A/a · Cream N/Cr · Gray variant class N/Nbuckskin · no modeled gray variant1/8 · 12.5%
Red factor E/r · Agouti A/a · Cream Cr/Cr · Gray variant class N/Nperlino · no modeled gray variant1/16 · 6.25%
Red factor E/r · Agouti A/A · Cream N/N · Gray variant class N/Nbay · no modeled gray variant1/32 · 3.125%
Red factor E/r · Agouti A/A · Cream N/Cr · Gray variant class N/Nbuckskin · no modeled gray variant1/16 · 6.25%
Red factor E/r · Agouti A/A · Cream Cr/Cr · Gray variant class N/Nperlino · no modeled gray variant1/32 · 3.125%
Red factor E/E · Agouti a/a · Cream N/N · Gray variant class N/Nblack · no modeled gray variant1/64 · 1.5625%
Red factor E/E · Agouti a/a · Cream N/Cr · Gray variant class N/Nsmoky black · no modeled gray variant1/32 · 3.125%
Red factor E/E · Agouti a/a · Cream Cr/Cr · Gray variant class N/Nsmoky cream · no modeled gray variant1/64 · 1.5625%
Red factor E/E · Agouti A/a · Cream N/N · Gray variant class N/Nbay · no modeled gray variant1/32 · 3.125%
Red factor E/E · Agouti A/a · Cream N/Cr · Gray variant class N/Nbuckskin · no modeled gray variant1/16 · 6.25%
Red factor E/E · Agouti A/a · Cream Cr/Cr · Gray variant class N/Nperlino · no modeled gray variant1/32 · 3.125%
Red factor E/E · Agouti A/A · Cream N/N · Gray variant class N/Nbay · no modeled gray variant1/64 · 1.5625%
Red factor E/E · Agouti A/A · Cream N/Cr · Gray variant class N/Nbuckskin · no modeled gray variant1/32 · 3.125%
Red factor E/E · Agouti A/A · Cream Cr/Cr · Gray variant class N/Nperlino · no modeled gray variant1/64 · 1.5625%

Export the assumptions and evidence

CSV includes every possible parental pairing’s coat probabilities, bounds and their witnesses, plus the selected pairing’s gametes, cells and complete genotype ledger. Counts use exact denominators of 256 for offspring and 16 for gametes.

How to use this calculator

  1. Load Known carrier pairing or Uncertain bay genotypes. These examples have explicit assumptions for every locus; they are not an inference from a photograph or breed.
  2. Open Edit parental genotypes. For both parents, choose the known genotype or every plausible candidate at red factor, agouti, cream and gray. Use Unknown explicitly rather than leaving a required choice blank. Update inheritance.
  3. Select one or more target color/gray classes. Read the selected pairing’s exact probability separately from the range across all candidate pairings. Inspect a target minimum or maximum to see the pairing that produces it.
  4. Choose an individual genotype for each parent to inspect. Compare its full probability distribution with the uncertainty intervals, then select a gamete-grid cell and trace its inherited genotype and probability.
  5. Review the offspring genotype ledger. Download inheritance CSV or use the selectable copy to retain inputs, all alternative pairings, target bounds, endpoint witnesses and selected-pairing details.

The Formula

The mathematical logic powering this calculation is:

At each independent locus, each parent contributes one of its two alleles with equal transmission probability. Encode the number of modeled alleles as p and q, each 0, 1 or 2. Offspring copy-count weights over denominator 4 are w0=(2−p)(2−q), w2=pq, w1=4−w0−w2. Multiply the four locus weights to obtain 81 offspring genotype classes with integer weights over denominator 256. Identical gametes are combined; their weights use denominator 16, and each gamete-grid cell multiplies its two header probabilities.
Red factor counts functional E alleles; r groups recessive red-factor alleles rather than distinguishing mutation-specific laboratory results. Zero E copies produces a chestnut base regardless of agouti. With E present, zero A copies produces a black base and one or two A copies a bay base. Cream copies 0/1/2 map chestnut to chestnut/palomino/cremello, bay to bay/buckskin/perlino, and black to black/smoky black/smoky cream. Gray absence or presence is retained alongside that underlying color class. G groups gray alleles such as G2 and G3 only for inheritance presence; this model does not distinguish their specific sequences or predict speed of graying, shade or health.
An entered set of candidate genotypes is not a probability distribution. The Cartesian product gives at most 81 genotypes per parent and 6,561 parental pairings. For each outcome, take the minimum and maximum probability across these complete pairings, retaining the actual pairing that attains each endpoint. Different outcomes can attain their endpoints in different pairings; the bounds cannot be added as a distribution. For a multi-outcome target, sum the exclusive outcome probabilities inside each pairing before finding its bounds. An empty target is zero; a target containing all classes is one.
The selected pairing is a member of the candidate set, not an inferred most-likely genotype or weighted average. Equal transmission within a known genotype does not imply equally likely candidate parental genotypes. Every finite candidate pairing is evaluated, without a simulation or search cap. Fractions and percentages are exact for the modeled denominators; CSV preserves every scenario’s integer weights.
Assumptions: independent allele classes, equal Mendelian transmission, and no mutation, penetrance or viability selection. No dun, silver, champagne, pearl, mushroom, roan, white spotting, dominant white, leopard complex, breed-frequency prior, linkage model, health assessment or parentage inference. Other genes can alter actual appearance. These predictions are conditional on the modeled allele classes and entered candidate sets, not a guarantee of coat appearance or breeding suitability.

Worked Example

If you input:

  • Parents: Both E/r, A/a, N/Cr, N/N at gray; no other color modifiers modeled.
  • Target: Buckskin with no modeled gray variant.

Result:

Each parent has eight equally likely gametes. The base-color probabilities are chestnut 1/4, bay 9/16 and black 3/16. Cream gives a 1/2 chance of one Cr allele, so buckskin has probability (9/16) × (1/2) = 9/32 = 28.125%. The other modeled non-gray classes are chestnut 6.25%, palomino 12.5%, cremello 6.25%, bay 14.0625%, perlino 14.0625%, black 4.6875%, smoky black 9.375% and smoky cream 4.6875%. Both gray genotypes are N/N, so modeled gray presence is zero. These probabilities describe the entered genotype assumptions, not guaranteed counts among foals.

Frequently Asked Questions

Can two bay horses produce a chestnut foal?

Under this model, yes if both carry a recessive red-factor allele. Two E/r parents give a 1/4 chance of r/r, which masks agouti and produces a chestnut base. If either parent is E/E, the modeled chance of a chestnut base is zero.

Does Unknown mean every genotype is equally likely?

No. Unknown includes all three genotype classes as possibilities without assigning them weights. The output is a range across complete parental pairings. A frequency-weighted average would need additional justified information that this planner does not assume.

Why do the maximum probabilities add to more than 100%?

Each maximum may come from a different pairing. For uncertain bay parental genotypes, bay can range from 56.25% to 100%, chestnut from 0% to 25%, and black from 0% to 25%. Those upper bounds are not one distribution. Select a union of outcomes to obtain its valid combined bounds.

Is a gray outcome the coat color at birth?

No. Gray causes progressive change and can act over different underlying colors. The planner retains the modeled base/dilution class and reports gray-variant inheritance separately. It does not predict a foal’s appearance at a particular age.

What do r and G mean in the input labels?

They are modeled allele classes. r groups recessive red-factor alleles; G groups gray variants such as G2/G3 only for presence inheritance. They are not mutation-specific test results, and the model does not estimate variant-specific effects or health.

Why does cream produce different colors on different bases?

The phenotype mapping depends on both base pigment and cream dosage. One cream copy gives palomino on chestnut, buckskin on bay, or smoky black on black; two copies give cremello, perlino or smoky cream respectively. Other modifiers are outside this four-locus model.

Can I infer the parents’ complete genotypes from their appearance?

This planner does not do that. Appearance may hide genotype information, and other genes can modify it. Enter known results or explicit candidate sets. Presets state their own illustrative assumptions rather than silently inferring missing loci.

Does this replace a full coat panel or assess breeding suitability?

No. Numerous dilution, pattern and other variants are omitted. The model does not assess health, genetic diversity, viability, parentage or breeding fitness. It is an inspectable inheritance calculation conditional on four entered allele classes.

Similar Calculators