Calculate epistasis ratios from an AaBb × AaBb cross and see how gene interaction changes the classic 9:3:3:1 pattern. The tool covers 9:3:4, 12:3:1, 15:1, 9:7, 13:3, and 9:6:1 ratios with live expected counts, phenotype grouping, and chi-square comparison.
Select a gene interaction model, enter observed offspring counts, and compare your data with the expected modified dihybrid ratio.
Each preset starts with an AaBb × AaBb cross, then groups the 16 Punnett boxes by the selected gene interaction.
Select the modified dihybrid ratio that matches the biological relationship between the two genes.
Recessive epistasis
A homozygous recessive genotype at one locus masks the second locus.
Expected F₂ ratio: 9:3:4
Classic example: Labrador coat colour logic, where ee blocks pigment deposition.
Use real offspring counts to test whether your data fit the selected epistasis ratio.
Live epistasis result
The largest expected phenotype class is A_B_ phenotype at 56.25%. Your observed counts fit this model at α = 0.05.
χ²
0
df
2
p-value
1.000
Each AaBb × AaBb genotype box keeps its genotype, while the colour shows the phenotype class after gene interaction.
AABB
A_B_
AABb
A_B_
AaBB
A_B_
AaBb
A_B_
AABb
A_B_
AAbb
A_bb
AaBb
A_B_
Aabb
A_bb
AaBB
A_B_
AaBb
A_B_
aaBB
aa__
aaBb
aa__
AaBb
A_B_
Aabb
A_bb
aaBb
aa__
aabb
aa__
The ratio values convert directly into expected probabilities and offspring counts.
Both loci carry a dominant allele, so the downstream phenotype appears.
56.25%
ratio 9
The A locus remains active, but bb changes the second trait state.
18.75%
ratio 3
The aa genotype hides the B locus, so aaB_ and aabb combine.
25.00%
ratio 4
| Phenotype class | Ratio | Probability | Observed | Expected | O − E | χ² contribution |
|---|---|---|---|---|---|---|
| A_B_ phenotype | 9 | 56.25% | 90 | 90 | 0 | 0 |
| A_bb phenotype | 3 | 18.75% | 30 | 30 | 0 | 0 |
| aa masked phenotype | 4 | 25.00% | 40 | 40 | 0 | 0 |
An epistasis Punnett square calculator explains why some two-gene crosses do not produce the standard 9:3:3:1 F2 ratio. It begins with the same AaBb × AaBb genotype grid, then collapses genotype boxes when one gene masks or modifies another gene. Students use it when textbook ratios such as 9:3:4 or 9:7 appear without a clear visual map.
William Bateson used the term epistasis in the early twentieth century while studying gene interactions that changed visible inheritance patterns. Modern genetics treats epistasis as interaction between loci, proteins, or pathway steps. Phillips reviewed this concept in 2008 and described epistasis as a central feature of genetic architecture. Read the Phillips epistasis review.
Epistasis differs from dominance because dominance compares alleles at one locus. Epistasis compares effects across loci. A recessive aa genotype can hide B/b, while a dominant A allele can suppress the second locus in a different model.
Choose a model such as recessive epistasis 9:3:4, dominant epistasis 12:3:1, complementary genes 9:7, or duplicate dominant epistasis 15:1.
Check how the 16 genotype boxes group into phenotype classes under the selected gene interaction.
Type the counted offspring number for each visible phenotype class rather than using percentages.
Use the probability cards, expected counts, and chi-square p-value to decide whether the selected ratio fits your data.
Use phenotype counts that match the selected model. For a 9:3:4 recessive epistasis test, enter counts for A_B_, A_bb, and aa__ rather than four separate 9:3:3:1 classes.
These buttons load common modified dihybrid ratios. Each preset changes how the 16 AaBb × AaBb genotype boxes collapse into phenotype groups.
This card defines the selected model, names the expected F2 ratio, and explains which gene masks, complements, or suppresses the other gene.
These inputs let you test real phenotype counts. The tool converts the selected ratio into expected counts and updates χ² without a calculate button.
The coloured grid keeps every genotype visible, then labels the epistatic phenotype group. This helps students see why classes merge.
The cards show ratio values, percentages, and expected counts. The table shows observed minus expected values, so you can identify which phenotype class drives the deviation.
A standard dihybrid cross produces 9:3:3:1 when two genes assort independently and affect separate traits. OpenStax explains this independent assortment model through the laws of inheritance. Review the OpenStax inheritance section.
Epistasis keeps independent assortment but changes phenotype interpretation. Recessive epistasis gives 9:3:4 when a homozygous recessive genotype masks another locus. Dominant epistasis gives 12:3:1 when one dominant allele blocks expression downstream.
Duplicate dominant epistasis gives 15:1 because either gene can produce the dominant phenotype. Complementary genes give 9:7 because both genes need a dominant allele. Dominant and recessive epistasis gives 13:3 when one dominant genotype class and the double recessive class share the same phenotype.
| Epistasis type | Modified ratio | Class grouping | Biological meaning |
|---|---|---|---|
| Recessive epistasis | 9:3:4 | A_B_, A_bb, aa__ | aa masks the second locus |
| Dominant epistasis | 12:3:1 | A___, aaB_, aabb | A masks B/b |
| Duplicate dominant | 15:1 | A___ or __B_, aabb | Either dominant gene works |
| Complementary genes | 9:7 | A_B_, all others | Both dominant loci are required |
| Dominant and recessive | 13:3 | A___ or aabb, aaB_ | Suppression and double loss merge |
A researcher counts 90 A_B_, 29 A_bb, and 41 aa__ offspring from an AaBb × AaBb cross. The total equals 160. A 9:3:4 model predicts 90, 30, and 40 offspring.
The deviations equal 0, −1, and +1. χ² stays near 0.06 with 2 degrees of freedom, so the data support the 9:3:4 model at α = 0.05.
A pigment pathway gives 92 coloured offspring and 68 colourless offspring. A 9:7 ratio with 160 total offspring predicts 90 coloured and 70 colourless offspring.
This pattern fits two required genes in the same pathway. A recessive block at either locus prevents the final pigment, so A_bb, aaB_, and aabb merge.
Epistasis helps explain why two traits can fail to behave like separate Mendelian characters. Plant breeders use this idea when pigment, disease resistance, seed traits, or flowering time depend on interacting pathway genes. A modified ratio can point toward enzyme order, suppressor genes, or redundant genes.
Human genetics also uses epistasis as a model for gene-gene interaction, especially in complex traits. A single-locus calculator cannot capture those relationships. This page stays focused on two-locus classroom crosses, but the same logic prepares students for pathway genetics and quantitative trait analysis.
A good ratio fit does not prove the molecular pathway. Several models can produce similar phenotype counts when sample size stays small. Linkage, incomplete penetrance, environmental effects, and scorer bias can also shift F2 classes.
Rare classes need enough expected offspring. A 15:1 ratio expects 1 recessive phenotype per 16 offspring, so a small family or seed tray may miss it by chance. Use follow-up crosses, molecular markers, or enzyme assays when the biological conclusion matters.
This calculator supports genetics education and experiment planning. It does not replace professional genetic counselling, diagnostic testing, or validated breeding decisions.
Use these calculators to compare standard Mendelian expectations with modified gene interaction ratios.
Calculate standard 9:3:3:1 offspring probabilities before comparing modified epistasis ratios.
Open CalculatorTest observed offspring counts against expected Mendelian or modified dihybrid ratios.
Open Calculator