Hardy–Weinberg principle

1 Weinberg, who first demonstrated it mathematically.

aa q 2 Aa 2pq AA p 2
0.8
1 Derivation
0.6
Consider a population of monoecious diploids, where
0.4 each organism produces male and female gametes at
equal frequency, and has two alleles at each gene locus.
Organisms reproduce by random union of gametes (the
0.2 “gene pool” population model). A locus in this population
has two alleles, A and a, that occur with initial frequen-
0 cies f 0 (A) = p and f 0 (a) = q, respectively.[1] The allele
p 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 frequencies at each generation are obtained by pooling
q 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
together the alleles from each genotype of the same gen-
eration according to the expected contribution from the
Hardy–Weinberg proportions for two alleles: the horizontal axis
homozygote and heterozygote genotypes, which are 1 and
shows the two allele frequencies p and q and the vertical axis
shows the expected genotype frequencies. Each line shows one 1/2, respectively:
of the three possible genotypes.

The Hardy–Weinberg principle, also known as the

Hardy–Weinberg equilibrium, model, theorem, or
law, states that allele and genotype frequencies in a popu-
lation will remain constant from generation to generation
in the absence of other evolutionary influences. These in-
fluences include mate choice, mutation, selection, genetic
drift, gene flow and meiotic drive. Because one or more of
these influences are typically present in real populations,
the Hardy–Weinberg principle describes an ideal condi-
tion against which the effects of these influences can be
analyzed.
In the simplest case of a single locus with two alleles
denoted A and a with frequencies f(A) = p and f(a) =
q, respectively, the expected genotype frequencies are
f(AA) = p2 for the AA homozygotes, f(aa) = q2 for the
aa homozygotes, and f(Aa) = 2pq for the heterozygotes.
The genotype proportions p2 , 2pq, and q2 are called the
Hardy–Weinberg proportions. Note that the sum of all
genotype frequencies of this case is the binomial expan-
sion of the square of the sum of p and q, and such a sum,
as it represents the total of all possibilities, must be equal
to 1. Therefore, (p + q)2 = p2 + 2pq + q2 = 1. The real- Length of p, q corresponds to allele frequencies (here p = 0.6, q
istic solution of this equation is q = 1 − p. = 0.4). Then area of rectangle represents genotype frequencies
(thus AA : Aa : aa = 0.36 : 0.48 : 0.16).
If union of gametes to produce the next generation is ran-
dom, it can be shown that the new frequency f′ satisfies
The different ways to form genotypes for the next gener-
f ′ (A) = f (A) and f ′ (a) = f (a) . That is, allele frequen-
ation can be shown in a Punnett square, where the pro-
cies are constant between generations, so equilibrium is
portion of each genotype is equal to the product of the
reached. row and column allele frequencies from the current gen-
The principle is named after G. H. Hardy and Wilhelm eration.

1
2 2 DEVIATIONS FROM HARDY–WEINBERG EQUILIBRIUM

The sum of the entries is p2 + 2pq + q2 = 1, as the geno- As before, one can show that the allele frequencies at time
type frequencies must sum to one. t+1 equal those at time t, and so, are constant in time.
Note again that as p + q = 1, the binomial expansion of (p Similarly, the genotype frequencies depend only on the
+ q)2 = p2 + 2pq + q2 = 1 gives the same relationships. allele frequencies, and so, after time t=1 are also constant
in time.
Summing the elements of the Punnett square or the bi-
nomial expansion, we obtain the expected genotype pro- If in either monoecious or dioecious organisms, either
the allele or genotype proportions are initially unequal
portions among the offspring after a single generation:
in either sex, it can be shown that constant proportions
are obtained after one generation of random mating. If
dioecious organisms are heterogametic and the gene locus
is located on the X chromosome, it can be shown that if
the allele frequencies are initially unequal in the two sexes
[e.g., XX females and XY males, as in humans], f′(a) in
the heterogametic sex ‘chases’ f(a) in the homogametic
These frequencies define the Hardy–Weinberg equilib- sex of the previous generation, until an equilibrium is
rium. It should be mentioned that the genotype frequen- reached at the weighted average of the two initial frequen-
cies after the first generation need not equal the genotype cies.
frequencies from the initial generation, e.g. f 1 (AA) ≠
f 0 (AA). However, the genotype frequencies for all fu-
ture times will equal the Hardy–Weinberg frequencies,
e.g. ft(AA) = f 1 (AA) for t > 1. This follows since the
2 Deviations from Hardy–
genotype frequencies of the next generation depend only Weinberg equilibrium
on the allele frequencies of the current generation which,
as calculated by equations (1) and (2), are preserved from The seven assumptions underlying Hardy–Weinberg
the initial generation: equilibrium are as follows:[3]

f1 (A) = f1 (AA) + 12 f1 (Aa) = p2 + pq = p(p + q) = p = •f0organisms

(A) are diploid
f1 (a) = f1 (aa) + 21 f1 (Aa) = q 2 + pq = q(p + q) = q = f•0 (a)
only sexual reproduction occurs
For the more general case of dioecious diploids [organ-
isms are either male or female] that reproduce by ran- • generations are non overlapping
dom mating of individuals, it is necessary to calculate the
• mating is random
genotype frequencies from the nine possible matings be-
tween each parental genotype (AA, Aa, and aa) in either • population size is infinitely large
sex, weighted by the expected genotype contributions of
each such mating.[2] Equivalently, one considers the six • allele frequencies are equal in the sexes
unique diploid-diploid combinations:
• there is no migration, mutation or selection

[(AA, AA), (AA, Aa), (AA, aa), (Aa, Aa), (Aa, aa), (aa, aa)] Violations of the Hardy–Weinberg assumptions can cause
and constructs a Punnett square for each, so as to calcu- deviations from expectation. How this affects the popu-
late its contribution to the next generation’s genotypes. lation depends on the assumptions that are violated.
These contributions are weighted according to the prob-
ability of each diploid-diploid combination, which fol- • Random mating. The HWP states the population
lows a multinomial distribution with k = 3. For example, will have the given genotypic frequencies (called
the probability of the mating combination (AA,aa) is 2 Hardy–Weinberg proportions) after a single genera-
ft(AA)ft(aa) and it can only result in the Aa genotype: tion of random mating within the population. When
[0,1,0]. Overall, the resulting genotype frequencies are the random mating assumption is violated, the popu-
calculated as: lation will not have Hardy–Weinberg proportions. A
common cause of non-random mating is inbreeding,
which causes an increase in homozygosity for all
[ft+1 (AA), ft+1 (Aa), ft+1 (aa)] =
[ ] genes.
= ft (AA)ft (AA) [1, 0, 0] + 2ft (AA)ft (Aa) 12 , 12 , 0 + 2ft (AA)ft (aa) [0, 1, 0]
[ ] [ ]
+ ft (Aa)ft (Aa) 41 , 12 , 14 + 2ft (Aa)ft (aa) 0, If12 , a12 population
+ ft (aa)ftviolates one
(aa) [0, 0, 1] of the following four as-
[( ) ( ) (
sumptions, the ) (
population may )2 ]have Hardy–
continue to
2
= ft (AA) + 21 ft (Aa) , 2 ft (AA) + 12 ft (Aa) ft (aa) + 12 ft (Aa) , ft (aa) + 21 ft (Aa)
Weinberg proportions each generation, but the allele fre-
[ ]
= ft (A)2 , 2ft (A)ft (a), ft (a)2 quencies will change over time.
4.1 Generalization for more than two alleles 3

• Selection, in general, causes allele frequencies to 4.1 Generalization for more than two alle-
change, often quite rapidly. While directional selec- les
tion eventually leads to the loss of all alleles except
the favored one, some forms of selection, such as
balancing selection, lead to equilibrium without loss
of alleles.

• Mutation will have a very subtle effect on allele fre-

quencies. Mutation rates are of the order 10−4 to
10−8 , and the change in allele frequency will be,
at most, the same order. Recurrent mutation will
maintain alleles in the population, even if there is
strong selection against them.

• Migration genetically links two or more populations
together. In general, allele frequencies will become
more homogeneous among the populations. Some
models for migration inherently include nonrandom
mating (Wahlund effect, for example). For those
models, the Hardy–Weinberg proportions will nor-
mally not be valid.
Punnett square for three-allele case (left) and four-allele case
(right). White areas are homozygotes. Colored areas are het-
erozygotes.
Consider an extra allele frequency, r. The two-allele case
is the binomial expansion of (p + q)2 , and thus the three-
allele case is the trinomial expansion of (p + q+ r)2 .

• Small population size can cause a random change

in allele frequencies. This is due to a sampling ef- (p + q + r)2 = p2 + q 2 + r2 + 2pq + 2pr + 2qr
fect, and is called genetic drift. Sampling effects are
most important when the allele is present in a small More generally, consider the alleles A1 , ..., An given by
number of copies. the allele frequencies p1 to pn;

3 Sex linkage (p1 + · · · + pn )2

giving for all homozygotes:

Where the A gene is sex linked, the heterogametic sex
(e.g., mammalian males; avian females) have only one
copy of the gene (and are termed hemizygous), while the
f (Ai Ai ) = p2i
homogametic sex (e.g., human females) have two copies.
The genotype frequencies at equilibrium are p and q for and for all heterozygotes:
the heterogametic sex but p2 , 2pq and q2 for the homoga-
metic sex.
For example, in humans red–green colorblindness is an f (Ai Aj ) = 2pi pj
X-linked recessive trait. In western European males, the
trait affects about 1 in 12, (q = 0.083) whereas it affects
about 1 in 200 females (0.005, compared to q2 = 0.007), 4.2 Generalization for polyploidy
very close to Hardy–Weinberg proportions.
The Hardy–Weinberg principle may also be generalized
If a population is brought together with males and females
to polyploid systems, that is, for organisms that have more
with a different allele frequency in each subpopulation
than two copies of each chromosome. Consider again
(males or females), the allele frequency of the male popu-
only two alleles. The diploid case is the binomial expan-
lation in the next generation will follow that of the female
sion of:
population because each son receives its X chromosome
from its mother. The population converges on equilib-
rium very quickly.
(p + q)2

and therefore the polyploid case is the polynomial expan-

4 Generalizations sion of:

The simple derivation above can be generalized for more

than two alleles and polyploidy. (p + q)c
4 6 SIGNIFICANCE TESTS FOR DEVIATION

where c is the ploidy, for example with tetraploid (c = 4):

assumption of the chi-squared distribution, will no longer
Depending on whether the organism is a 'true' tetraploid hold, and it may be necessary to use a form of Fisher’s
or an amphidiploid will determine how long it will take exact test, which requires a computer to solve. More re-
for the population to reach Hardy–Weinberg equilibrium. cently a number of MCMC methods of testing for devia-
tions from HWP have been proposed (Guo & Thompson,
1992; Wigginton et al. 2005)
4.3 Complete generalization
For n distinct alleles in c -ploids, the genotype frequencies
6.1 Example χ2 test for deviation
in the Hardy–Weinberg equilibrium are given by individ-
This data is from E.B. Ford (1971) on the Scarlet tiger
ual terms in the multinomial expansion of (p1 +· · ·+pn )c
moth, for which the phenotypes of a sample of the pop-
:
ulation were recorded. Genotype-phenotype distinction
is assumed to be negligibly small. The null hypothesis
∑ ( ) is that the population is in Hardy–Weinberg proportions,
c
(p1 +· · ·+pn )c = · · ·the
pk11and pknnalternative hypothesis is that the population is not
k1 , . . . , k n in Hardy–Weinberg proportions.
k1 ,...,kn ∈N:k1 +···+kn =c

From this, allele frequencies can be calculated:

5 Applications
2 × obs(AA) + obs(Aa)
p=
The Hardy–Weinberg principle may be applied in two 2 × (obs(AA) + obs(Aa) + obs(aa))
ways, either a population is assumed to be in Hardy–
Weinberg proportions, in which the genotype frequencies
1469 × 2 + 138
can be calculated, or if the genotype frequencies of all =
three genotypes are known, they can be tested for devia- 2 × (1469 + 138 + 5)
tions that are statistically significant.
3076
=
3224
5.1 Application to cases of complete domi-
nance = 0.954
Suppose that the phenotypes of AA and Aa are indistin- and
guishable, i.e., there is complete dominance. Assuming
that the Hardy–Weinberg principle applies to the popula-
tion, then q can still be calculated from f(aa): q =1−p
= 1 − 0.954
√ = 0.046
q = f (aa)
So the Hardy–Weinberg expectation is:
and p can be calculated from q . And thus an estimate
of f(AA) and f(Aa) derived from p2 and 2pq respec-
Exp(AA) = p2 n = 0.9542 × 1612 = 1467.4
tively. Note however, such a population cannot be tested
for equilibrium using the significance tests below because Exp(Aa) = 2pqn = 2 × 0.954 × 0.046 × 1612 = 141.2
it is assumed a priori. Exp(aa) = q 2 n = 0.0462 × 1612 = 3.4
Pearson’s chi-squared test states:
6 Significance tests for deviation
∑ (O − E)2
χ2 =
Testing deviation from the HWP is generally performed E
using Pearson’s chi-squared test, using the observed geno- (1469 − 1467.4)2 (138 − 141.2)2 (5 − 3.4)2
type frequencies obtained from the data and the expected = + +
1467.4 141.2 3.4
genotype frequencies obtained using the HWP. For sys- = 0.001 + 0.073 + 0.756
tems where there are large numbers of alleles, this may re-
= 0.83
sult in data with many empty possible genotypes and low
genotype counts, because there are often not enough in- There is 1 degree of freedom (degrees of freedom for test
dividuals present in the sample to adequately represent all for Hardy–Weinberg proportions are # genotypes − # al-
genotype classes. If this is the case, then the asymptotic leles). The 5% significance level for 1 degree of freedom
 5

is 3.84, and since the χ2 value is less than this, the null For two alleles, the chi-squared goodness of fit test for
hypothesis that the population is in Hardy–Weinberg fre- Hardy–Weinberg proportions is equivalent to the test for
quencies is not rejected. inbreeding, F = 0.
The inbreeding coefficient is unstable as the expected
value approaches zero, and thus not useful for rare and
6.2 Fisher’s exact test (probability test)
very common alleles. For: E = 0, O > 0, F = −∞ and E
= 0, O = 0, F is undefined.
Fisher’s exact test can be applied to testing for Hardy–
Weinberg proportions. Since the test is conditional on the
allele frequencies, p and q, the problem can be viewed as
testing for the proper number of heterozygotes. In this 8 History
way, the hypothesis of Hardy–Weinberg proportions is
rejected if the number of heterozygotes is too large or too Mendelian genetics were rediscovered in 1900. However,
small. The conditional probabilities for the heterozygote, it remained somewhat controversial for several years as it
given the allele frequencies are given in Emigh (1980) as was not then known how it could cause continuous char-
acteristics. Udny Yule (1902) argued against Mendelism
because he thought that dominant alleles would increase
( )
n in the population.[5] The American William E. Castle
n11 ,n12 ,n22
prob[n12 |n1 ] = ( 2n ) 2 , n12
(1903) showed that without selection, the genotype fre-
n1 ,n2 quencies would remain stable.[6] Karl Pearson (1903)
found one equilibrium position with values of p = q
where n11 , n12 , n22 are the observed numbers of the three
= 0.5.[7] Reginald Punnett, unable to counter Yule’s
genotypes, AA, Aa, and aa, respectively, and n1 is the
point, introduced the problem to G. H. Hardy, a British
number of A alleles, where n1 = 2n11 + n12 .
mathematician, with whom he played cricket. Hardy was
An example Using one of the examples from Emigh a pure mathematician and held applied mathematics in
(1980),[4] we can consider the case where n = 100, and some contempt; his view of biologists’ use of mathemat-
p = 0.34. The possible observed heterozygotes and their ics comes across in his 1908 paper where he describes
exact significance level is given in Table 4. this as “very simple":[8]
Using this table, one must look up the significance level
of the test based on the observed number of heterozy- To the Editor of Science: I am reluctant to
gotes. For example, if one observed 20 heterozygotes, intrude in a discussion concerning matters of
the significance level for the test is 0.007. As is typical which I have no expert knowledge, and I should
for Fisher’s exact test for small samples, the gradation of have expected the very simple point which I
significance levels is quite coarse. wish to make to have been familiar to biologists.
However, some remarks of Mr. Udny Yule, to
However, a table like this has to be created for every ex- which Mr. R. C. Punnett has called my atten-
periment, since the tables are dependent on both n and tion, suggest that it may still be worth making...
p.
Suppose that Aa is a pair of Mendelian char-
acters, A being dominant, and that in any given
7 Inbreeding coefficient generation the number of pure dominants (AA),
heterozygotes (Aa), and pure recessives (aa)
are as p:2q:r. Finally, suppose that the num-
The inbreeding coefficient, F (see also F-statistics), is one
bers are fairly large, so that mating may be re-
minus the observed frequency of heterozygotes over that
garded as random, that the sexes are evenly dis-
expected from Hardy–Weinberg equilibrium.
tributed among the three varieties, and that all
are equally fertile. A little mathematics of the
E (f (Aa))−O(f (Aa)) O(f (Aa))
F = E(f (Aa)) =1− E(f (Aa)) , multiplication-table type is enough to show that
in the next generation the numbers will be as (p
where the expected value from Hardy–Weinberg equilib- + q)2 :2(p + q)(q + r):(q + r)2 , or as p1 :2q1 :r1 ,
rium is given by say.

The interesting question is: in what circum-

stances will this distribution be the same as that
E(f (Aa)) = 2pq in the generation before? It is easy to see that
the condition for this is q2 = pr. And since q1 2
For example, for Ford’s data above; = p1 r1 , whatever the values of p, q, and r may
be, the distribution will in any case continue un-
F =1− 138
141.2 = 0.023. changed after the second generation
6 8 HISTORY

The principle was thus known as Hardy’s law in the
English-speaking world until 1943, when Curt Stern
pointed out that it had first been formulated indepen-
dently in 1908 by the German physician Wilhelm Wein-
berg.[9][10] William Castle in 1903 also derived the ra-
tios for the special case of equal allele frequencies, and
it is sometimes (but rarely) called the Hardy–Weinberg–
Castle Law.
Where the final equality holds because the allele propor-
2
tions must sum to one. In both cases, qt−1 = pt−1 rt−1
. It can be shown that the other two equilibrium condi-
tions imply the same equation. Together, the solutions
of the three equilibrium equations imply sufficiency of
Hardy’s condition for equilibrium. Since the condition al-
ways holds for the second generation, all succeeding gen-
erations have the same proportions.

8.1 Derivation of Hardy’s equations 8.2 Numerical example

Hardy’s statement begins with a recurrence relation for An example computation of the genotype distribution
the frequencies p, 2q, and r. These recurrence relations given by Hardy’s original equations is instructive. The
follow from fundamental concepts in probability, specif- phenotype distribution from Table 3 above will be used
ically independence, and conditional probability. For ex- to compute Hardy’s initial genotype distribution. Note
ample, consider the probability of an offspring from the that the p and q values used by Hardy are not the same as
generation t being homozygous dominant. Alleles are in- those used above.
herited independently from each parent. A dominant al-
lele can be inherited from a homozygous dominant par-
ent with probability 1, or from a heterozygous parent with sum = obs(AA) + 2 × obs(Aa) + obs(aa) = 1469 + 2 × 138 + 5
probability 0.5. To represent this reasoning in an equa- = 1750
tion, let At represent inheritance of a dominant allele 1469
from a parent. Furthermore, let AAt−1 and Aat−1 rep- p= = 0.83943
1750
resent potential parental genotypes in the preceding gen- 2 × 138
eration. 2q = = 0.15771
1750
5
r= = 0.00286
2 1750
pt = P (At , At ) = P (At )
As checks on the distribution, compute
2
= (P (At |AAt−1 )P (AAt−1 ) + P (At |Aat−1 )P (Aat−1 ))
2
= ((1)pt−1 + (0.5)2qt−1 )
p + 2q + r = 0.83943 + 0.15771 + 0.00286 = 1.00000
2
= (pt−1 + qt−1 )
and
The same reasoning, applied to the other genotypes yields
the two remaining recurrence relations. Equilibrium oc-
curs when each proportion is constant between subse- E0 = q − pr = 0.00382.
2

quent generations. More formally, a population is at equi- For the next generation, Hardy’s equations give
librium at generation t when

0 = pt − pt−1 , 0 = qt − qt−1 , and 0 = 0.15771

q= = 0.07886
rt − rt−1 2

By solving these equations necessary and sufficient condi- p1 = (p + q)2 = 0.84325

tions for equilibrium to occur can be determined. Again,
2q1 = 2(p + q)(q + r) = 0.15007
consider the frequency of homozygous dominant animals.
Equilibrium implies r1 = (q + r)2 = 0.00668.
Again as checks on the distribution, compute
0 = pt − pt−1
= p2t−1 + 2pt−1 qt−1 + qt−1
2
− pt−1 p1 +2q1 +r1 = 0.84325+0.15007+0.00668 = 1.00000

First consider the case, where pt−1 = 0 , and note that it and
implies that qt−1 = 0 and rt−1 = 1 . Now consider the
remaining case, where pt−1 ≠ 0
E1 = q12 − p1 r1 = 0.00000
which are the expected values. The reader may demon-
2
0 = pt−1 (pt−1 + 2qt−1 + qt−1 /pt−1 − 1) strate that subsequent use of the second-generation values
2
= qt−1 /pt−1 − rt−1 for a third generation will yield identical results.

9 Graphical representation
A de Finetti diagram representing a distribution of genotype fre-
quencies
It is possible to represent the distribution of genotype fre-
quencies for a bi-allelic locus within a population graph-
ically using a de Finetti diagram. This uses a triangular
plot (also known as trilinear, triaxial or ternary plot) to
represent the distribution of the three genotype frequen-
cies in relation to each other. It differs from many other
such plots in that the direction of one of the axes has
been reversed.[11] The curved line in the diagram is the
Hardy–Weinberg parabola and represents the state where
alleles are in Hardy–Weinberg equilibrium. It is possible
to represent the effects of natural selection and its effect
on allele frequency on such graphs. [12] The de Finetti
diagram has been developed and used extensively by A.
W. F. Edwards in his book Foundations of Mathematical
Genetics.[13]
10 See also
• Regression toward the mean
• Multinomial distribution (Hardy–Weinberg is a tri-
nomial distribution with probabilities (θ2 , 2θ(1 −
θ), (1 − θ)2 ) )
11 References
[1] The term frequency usually refers to a number or count,
but in this context, it is synonymous with probability.
[2] http://www.mun.ca/biology/scarr/2900_HW_for_
dioecious.html
7
[3] Hartl DL, Clarke AG (2007) Principles of population ge-
netics. Sunderland, MA: Sinauer
[4] Emigh, Ted H. (1980). “A Comparison of Tests for
Hardy–Weinberg Equilibrium”. Biometrics. 36 (4): 627–
642. doi:10.2307/2556115. JSTOR 2556115.
[5] Yule, 1902
[6] Castle, 1903
[7] Pearson, 1903
[8] Hardy, 1908
[9] Crow, James F. (1999). “Hardy, Weinberg and language
impediments”. Genetics. 152 (3): 821–825. PMC
1460671 . PMID 10388804.
[10] Stern, Curt (1962). “Wilhelm Weinberg”. Genetics. 47:
1–5.
[11] Cannings, C.; Edwards, A.W.F. (1968). “Natural selec-
tion and the de Finetti diagram”. Annals of Human Ge-
netics. 31: 421–428.
[12] See e.g. Ineichen & Batschelet 1975
[13] Edwards, 1977
12 Bibliography
• Castle, W. E. (1903). “The laws of Galton and
Mendel and some laws governing race improvement
by selection”. Proceedings of the American Academy
of Arts and Sciences. 35: 233–242.
• Crow, Jf (Jul 1999). “Hardy, Weinberg and lan-
guage impediments”. Genetics. 152 (3): 821–
5. ISSN 0016-6731. PMC 1460671 . PMID
10388804.
• Edwards, A.W.F. 1977. Foundations of Mathemat-
ical Genetics. Cambridge University Press, Cam-
bridge (2nd ed., 2000). ISBN 0-521-77544-2
• Emigh, T.H. (1980). “A comparison of tests for
Hardy–Weinberg equilibrium”. Biometrics. 36 (4):
627–642. doi:10.2307/2556115. JSTOR 2556115.
• Ford, E.B. (1971). Ecological Genetics, London.
• Guo, Sw; Thompson, Ea (Jun 1992). “Per-
forming the exact test of Hardy–Weinberg pro-
portion for multiple alleles”. Biometrics. Bio-
metrics, Vol. 48, No. 2. 48 (2): 361–72.
doi:10.2307/2532296. ISSN 0006-341X. JSTOR
2532296. PMID 1637966.
• Hardy, G. H. (Jul 1908). “Mendelian Proportions in
a Mixed Population” (PDF). Science. 28 (706): 49–
50. doi:10.1126/science.28.706.49. ISSN 0036-
8075. PMID 17779291.
8 13 EXTERNAL LINKS

• Ineichen, Robert; Batschelet, Eduard (1975).

“Genetic selection and de Finetti diagrams”.
Journal of Mathematical Biology. 2: 33.
doi:10.1007/BF00276014.
• Masel, Joanna (2012). “Rethinking Hardy–
Weinberg and genetic drift in undergradu-
ate biology”. BioEssays. 34 (8): 701–10.
doi:10.1002/bies.201100178. PMID 22576789.
• Pearson, K. (1903). “Mathematical contributions
to the theory of evolution. XI. On the influ-
ence of natural selection on the variability and
correlation of organs”. Philosophical Transactions
of the Royal Society A. 200 (321–330): 1–66.
doi:10.1098/rsta.1903.0001.

• Stern, C. (1943). “The Hardy–Weinberg

law”. Science. 97 (2510): 137–138.
doi:10.1126/science.97.2510.137. JSTOR
1670409. PMID 17788516.

• Weinberg, W. (1908). "Über den Nachweis der

Vererbung beim Menschen”. Jahreshefte des Vere-
ins für vaterländische Naturkunde in Württemberg.
64: 368–382.

• Wigginton, Je; Cutler, Dj; Abecasis, Gr (May

2005). “A Note on Exact Tests of Hardy–Weinberg
Equilibrium”. American Journal of Human Genet-
ics. 76 (5): 887–93. doi:10.1086/429864. ISSN
0002-9297. PMC 1199378 . PMID 15789306.

• Yule, G. U. (1902). “Mendel’s laws and their prob-

able relation to intra-racial heredity”. New Phy-
tol. 1 (193–207): 222–238. doi:10.1111/j.1469-
8137.1902.tb07336.x.

13 External links
• EvolutionSolution (at bottom of page)

• Hardy–Weinberg Equilibrium Calculator

• genetics Population Genetics Simulator

• HARDY C implementation of Guo & Thompson

1992

• Source code (C/C++/Fortran/R) for Wigginton et al.

2005

• Online de Finetti Diagram Generator and Hardy–

Weinberg equilibrium tests

• Online Hardy–Weinberg equilibrium tests and

drawing of de Finetti diagrams

• Hardy–Weinberg Equilibrium Calculator

 9

14 Text and image sources, contributors, and licenses

14.1 Text
• Hardy–Weinberg principle Source: https://en.wikipedia.org/wiki/Hardy%E2%80%93Weinberg_principle?oldid=779723472 Contribu-
tors: AxelBoldt, Toby Bartels, Michael Hardy, Lexor, Sannse, Nikai, Thg, Charles Matthews, Grendelkhan, J D, Robbot, RedWolf, Hen-
rygb, Giftlite, Duncharris, Bobblewik, MarkSweep, Neffk, Rdsmith4, Thorwald, Kate, Discospinster, Rich Farmbrough, TedPavlic, Wiki-
acc, Dave souza, Imoen, Euyyn, Guettarda, Barbirossa, Moop, Mel Etitis, OwenX, LOL, Kzollman, JeremyA, GregorB, Tmetz, Btyner,
RichardWeiss, BD2412, Grammarbot, Rjwilmsi, MarSch, Salix alba, Fred Bradstadt, Mathbot, JYOuyang, Thecurran, Chobot, Flcelloguy,
YurikBot, Sceptre, RussBot, Mysid, DeadEyeArrow, Chefyingi, Isoxyl, SmackBot, Eperotao, Jim62sch, Bomac, TimBentley, VMS Mo-
saic, Khoikhoi, TedE, Infovoria, Richard001, Vina-iwbot~enwiki, SashatoBot, Acidburn24m, Joelmills, Extremophile, Xaura, Bendzh,
Dicklyon, Johnchiu, Chris55, Dycedarg, Lavateraguy, CBM, Metzenberg, Slack---line, Gregbard, Was a bee, Wikipediarules2221, En-
glishnerd, Talgalili, Thijs!bot, Headbomb, Openlander, Mentifisto, Thankful~enwiki, TimVickers, TuvicBot, UAucklandLibr, MartinBot,
STBot, Rettetast, Burnedthru, CommonsDelinker, Genetics411, Mausy5043, Dr d12, Ryan Postlethwaite, Chiswick Chap, Nwbeeson,
VolkovBot, DrMicro, A.Ou, Philip Trueman, TXiKiBoT, Red Act, Oxfordwang, SpecMode, Logan, SieBot, Nihil novi, Chrissen, Zbvhs,
Yerpo, Hello71, Melcombe, Huku-chan, ImageRemovalBot, ClueBot, GorillaWarfare, Theseven7, Cp111, MrKIA11, CPeterman, Excirial,
Variousbagels, L0mars01, Project FMF, Aitias, Amaltheus, Johnuniq, Kaktusz, XLinkBot, Addbot, Jacopo Werther, Rbcurrie, I am not
a dog, Download, Akita86, Tyw7, Kisbesbot, Tide rolls, Lightbot, Zorrobot, Swarm, Luckas-bot, Yobot, BruceMcAdam, AnomieBOT,
Haetmachine, Maynard Jones, Collieuk, Citation bot, Xqbot, Drilnoth, DSisyphBot, Plasmon1248, Riotrocket8676, RibotBOT, SassoBot,
Shadowjams, Smcewincarr, BadalJ, Citation bot 1, RockfangSemi, Serols, Fumitol, Aricooperdavis, Spritrig, Stroppolo, Reach Out to
the Truth, RjwilmsiBot, Kastchei, Yangosplat222, EmausBot, John of Reading, Lbui20, Da500063, PoeticVerse, Dcirovic, Tuxedo junc-
tion, Josve05a, Habstinat, Gz33, Seattle, Joannamasel, ClueBot NG, Mohanapriya94, Cntras, Helpful Pixie Bot, Plantdrew, BG19bot,
RobLandau, Tazerdadog, CitationCleanerBot, Duxwing, The Pikachu Who Dared, BattyBot, Barahona42, ‫محسن پورپونه‬, Makecat-bot,
Elise.flynn, FallingGravity, Ohl4xoaS, CsDix, Sosthenes12, Melias003, Comp.arch, Lanzhou1126, Anrnusna, Coz7, TuxLibNit, Monkbot,
HMSLavender, Truyopx, Latex-yow, Rileyskeengaar and Anonymous: 245

14.2 Images
• File:Commons-logo.svg Source: https://upload.wikimedia.org/wikipedia/en/4/4a/Commons-logo.svg License: PD Contributors: ? Origi-
nal artist: ?
• File:De_finetti_diagram.png Source: https://upload.wikimedia.org/wikipedia/commons/5/52/De_finetti_diagram.png License: Public
domain Contributors: Own work Original artist: Genisock2
• File:Hardy-Weinberg.svg Source: https://upload.wikimedia.org/wikipedia/commons/b/b2/Hardy-Weinberg.svg License: CC BY-SA
3.0 Contributors: Own work Original artist: Johnuniq
• File:Hardy–Weinberg_law_-_Punnett_square.svg Source: https://upload.wikimedia.org/wikipedia/commons/4/4d/Hardy%E2%80%
93Weinberg_law_-_Punnett_square.svg License: CC0 Contributors:
• File:Schemat punneta2.svg Original artist: own
• File:Hardy–Weinberg_law_-_Punnett_square2.svg Source: https://upload.wikimedia.org/wikipedia/commons/2/25/Hardy%E2%
80%93Weinberg_law_-_Punnett_square2.svg License: CC0 Contributors:
• File:Schemat punneta2.svg Original artist: own
• File:Lock-green.svg Source: https://upload.wikimedia.org/wikipedia/commons/6/65/Lock-green.svg License: CC0 Contributors: en:File:
Free-to-read_lock_75.svg Original artist: User:Trappist the monk

14.3 Content license

• Creative Commons Attribution-Share Alike 3.0