AP Biology · Unit 7 lecture · Part 6 of 12 · 8 min

7.5 Hardy-Weinberg equilibrium: calculationsDoing the math the AP way

Mrs. Otero
Hardy-Weinberg problems scare people, but there are only two ways in. The trick is knowing where to start.

Way one: you know the fraction with the recessive phenotype. Those individuals are aa, so that fraction is $q^2$. Take the square root to get q, subtract from 1 to get p, and the rest follows. Way two: you know how many individuals have each genotype, so you count alleles directly. Way two works even when the population is not in equilibrium.

?Wait, why can't I start from the dominant phenotype?
It includes two genotypes, AA ($p^2$) and Aa ($2pq$), so its frequency is $p^2 + 2pq$. Its square root isn't p. The recessive phenotype is only aa, so it is exactly $q^2$.
Way one: from the recessive phenotype
A recessive condition affects 1 in 2,500 newborns in a large population in equilibrium. Calculate (a) q, (b) p, (c) the frequency of carriers, and (d) about 1 in how many people is a carrier.
  1. (a) $q^2 = \frac{1}{2500} = 0.0004$, so $q = \sqrt{0.0004} = 0.02$.
    Why?
    Start from q², the only genotype you can see.
  2. (b) p = 1 − 0.02 = 0.98.
  3. (c) Carriers are Aa: 2pq = 2(0.98)(0.02) = 0.0392, about 3.9%.
  4. (d) 1 ÷ 0.0392 ≈ 25.5, so about 1 in 26 people.
    Why?
    1 in 2,500 affected, but 1 in 26 a carrier: almost 100 carriers per affected person.

Answer: (a) 0.02 (b) 0.98 (c) 0.0392 (d) about 1 in 26.

Stuck? Get a hint
  1. The affected fraction is q².
  2. √0.0004 = 0.02.
  3. Carriers are 2pq.
?Wait, how does a chi-square test work?
From Unit 5: $\chi^2 = \sum \frac{(o - e)^2}{e}$, with e from the null hypothesis. If $\chi^2$ is bigger than the critical value at p = 0.05, reject the null: the gap is too big to blame on chance.
Way two: count alleles, then test
200 deer mice: 150 AA, 40 Aa, 10 aa. (a) Find p and q. (b) Find the expected genotype numbers under Hardy-Weinberg. (c) Calculate chi-square and decide at p = 0.05. (d) Suggest a cause.
GenotypeObserved (o)Expected (e)(o - e)² / e
AA1500.7225 × 200 = 144.50.21
Aa400.255 × 200 = 51.02.37
aa100.0225 × 200 = 4.56.72
Total200200χ² = 9.30
  1. (a) A copies: 2(150) + 40 = 340 of 400 total, so p = 0.85 and q = 0.15.
    Why?
    $p = \frac{2(\text{AA}) + \text{Aa}}{2N}$. Each AA has 2 copies of A, each Aa has 1.
  2. (b) p² = 0.7225, 2pq = 0.255, q² = 0.0225. Times 200: 144.5 AA, 51.0 Aa, 4.5 aa.
    Why?
    Keep the decimals; don't round to whole mice.
  3. (c) χ² = 0.21 + 2.37 + 6.72 ≈ 9.30, which is greater than the critical value, so reject the null: not in equilibrium.
    Why?
    Many answer keys use 2 degrees of freedom here (5.99). Strictly a Hardy-Weinberg test has 1 (3.84), because p came from the same data. 9.30 beats both.
  4. (d) Too few heterozygotes and too many of both homozygotes fits nonrandom mating, like inbreeding.
    Why?
    Look at which way the data miss. Inbreeding makes homozygotes more common.

Answer: (a) p = 0.85, q = 0.15. (b) 144.5, 51.0, 4.5. (c) χ² ≈ 9.30, reject. (d) A heterozygote shortage, consistent with inbreeding.

Stuck? Get a hint
  1. Count A copies: 2 per AA and 1 per Aa, out of 400.
  2. Expected = p², 2pq and q² times 200.
  3. Add the three (o − e)²/e terms and compare.
observed150AA40Aa10aaexpected144.5AA51Aa4.5aa
Observed genotype counts next to the counts predicted from p = 0.85. Heterozygotes fall short; both homozygotes are above their predictions.
Pause and think
Inbreeding makes more AA and aa and fewer Aa. Does inbreeding by itself change p and q?
I thought about it. Show me
No. Mating relatives reshuffles alleles into more homozygotes, but no copies are added or removed. Nonrandom mating changes genotype frequencies; selection, drift, gene flow and mutation change allele frequencies.
Check yourself
In a large population, 36% of flowers are white (rr) and red is dominant. Assuming equilibrium, what percent are heterozygous?
  1. A16%
  2. B24%
  3. C48%
  4. D64%
Show the answer

C. q² = 0.36, so q = 0.6 and p = 0.4, and 2pq = 0.48. 64% is the trap: that's all red flowers (p² + 2pq).

Check yourself
Observed genotype counts differ from Hardy-Weinberg predictions, and chi-square is far above the critical value. Which conclusion is best supported?
  1. AThe population is in equilibrium.
  2. BAt least one Hardy-Weinberg condition is not being met for this gene.
  3. CNatural selection is definitely acting on this gene.
  4. DThe alleles must be codominant.
Show the answer

B. Rejecting the null means at least one condition is broken, but the test can't say which. 'Definitely selection' is the trap: drift, gene flow or nonrandom mating could also do it.

  • Recessive phenotype given: q² first, then q, then p = 1 − q.
  • Genotype counts given: $p = \frac{2(\text{AA}) + \text{Aa}}{2N}$, in or out of equilibrium.
  • Testing equilibrium: expected counts from p², 2pq, q², then chi-square. Rejecting means a condition is broken.
  • Nonrandom mating changes genotype frequencies, not p and q.