🧬 Mendelian Genetics

Pick a cross-type tab (Monohybrid, Dihybrid or Hardy–Weinberg) and a trait from the dropdown. For mono and dihybrid crosses, use the allele dropdowns to set each of Parent 1 and Parent 2's two alleles, then press "Run Cross". The square, phenotype colours and offspring ratios update automatically whenever you change an allele.

It performs a stochastic sampling: for each of 200 offspring it randomly picks one allele from each parent (a 50/50 coin flip per parent) and records the resulting phenotype. The bar chart of these counts shows how observed frequencies cluster around, but rarely land exactly on, the theoretical 3:1 or other ratio — illustrating sampling variation in real families and breeding experiments.

For a single gene with two alleles at frequencies p and q (where p + q = 1), the equilibrium genotype frequencies are p² for homozygous dominant, 2pq for heterozygous, and q² for homozygous recessive, and these sum to 1. The Hardy–Weinberg tab plots these three frequencies against p and converts them into approximate individual counts for your chosen population size N.

The equilibrium holds only for an idealised population with random mating, no mutation, no natural selection, no migration and infinitely large size (no genetic drift). Real populations violate these to some degree, so observed deviations from p², 2pq and q² are used as evidence that one of these evolutionary forces is acting on a gene.

The ABO trait uses three alleles: IA and IB are codominant with each other, while i is recessive to both. So IAIA and IAi give type A, IBIB and IBi give type B, IAIB gives type AB because both alleles are expressed, and ii gives type O. This shows that not every gene fits the simple two-allele dominant–recessive pattern.

The plant traits (flower colour and height) closely match Mendel's original pea experiments and are genuinely single-gene. The human eye-colour example is a deliberate simplification: real eye colour is polygenic, influenced by several genes, so the clean brown-over-blue model is a teaching device rather than a complete description.

The Punnett square gives expected probabilities, not guaranteed outcomes. Because fertilisation is random, a finite sample of 200 offspring will show statistical scatter around the ideal ratio. Larger samples converge closer to the predicted proportions, which is why geneticists rely on large counts and statistical tests rather than a handful of individuals.