Hardy-Weinberg Equilibrium Calculator

It is p squared plus 2pq plus q squared equals 1, where p is the dominant allele frequency and q is the recessive allele frequency. The three terms are the expected frequencies of the AA, Aa and aa genotypes when a population is in equilibrium at a single two-allele locus.

The model assumes p plus q equals 1, so knowing one allele frequency fixes the other. If you enter p it sets q to 1 minus p; if you enter q it sets p to 1 minus q. From those it computes AA as p squared, Aa as 2pq and aa as q squared.

Take the square root of the AA frequency to recover p, or the square root of the aa frequency to recover q. The other allele frequency is one minus that result. This works because AA equals p squared and aa equals q squared, so p equals sqrt(AA) and q equals sqrt(aa).

Because p plus q equals 1, squaring both sides expands (p + q) squared into p squared plus 2pq plus q squared, which therefore also equals 1. The three genotypes AA, Aa and aa are mutually exclusive and cover every individual, so their frequencies must total 1.

It assumes an infinitely large, randomly mating population with no mutation, no migration, no natural selection and no genetic drift. It also assumes non-overlapping generations and equal allele frequencies in both sexes. Real populations only approximate these conditions, so the equilibrium is a null model rather than an exact description.

The equilibrium is fully determined by a single allele or genotype frequency, so supplying more than one could be contradictory. The tool accepts exactly one input, and the underlying engine rejects a request that provides two or more values at once.

Enter every value as a decimal frequency between 0 and 1. A genotype seen in 16 percent of a population is 0.16, so type 0.16 rather than 16. Values outside 0 to 1 are rejected because a frequency cannot be negative or exceed the whole population.

This calculator gives the expected frequencies, not a statistical test. To decide whether real counts fit, compare the observed genotype numbers with these expected numbers using a chi-square goodness-of-fit test. A large deviation suggests selection, non-random mating or another force is acting on the locus.

Yes. The calculation is pure arithmetic performed locally in your browser, and no allele or genotype value is uploaded to a server. Nothing you enter leaves your device.