gENETIC dRIfT: EvolUTIoN AT RANdoM 175
The explanation is that much of the polymorphism in DNA within species
results from random genetic drift acting on selectively neutral mutations. On aver-
age, the heterozygosity resulting from neutral mutations evolving by drift in a dip-
loid species is expected to be π ≈ 4 Ne μn (7.1)where μn is the neutral mutation rate—that is, the chance per generation that the
locus mutates to another allele that does not change the organism’s fitness. For
example, if the total mutation rate at a locus is μ = 10–6 but only 10 percent of muta-
tions are selectively neutral, then the neutral mutation rate is μn = 0.1 × 10 –6 = 10–7.
Equation 7.1 represents the product of three quantities: the expected number of
generations back to the coalescence of two copies of a gene (2Ne); the probability
that a selectively neutral mutation occurs in a generation (μn); and a factor of 2 that
accounts for the fact that mutations could occur in either of the two lineages leading
back to their most recent common ancestor. (Equation 7.1 is an approximation that is
accurate when π is 0.1 or less, which is often the case.) To summarize, polymorphism
increases with the effective population size (Ne) and the neutral mutation rate (μn).
This simple result explains major patterns in genetic variation seen across the
genome. Most mutations that occur in coding regions of the genome are nonsyn-
onymous (they change an amino acid in the protein), and most changes to a protein
are deleterious (they decrease survival or reproduction) (see Chapter 4). These muta-
tions are weeded out from the population in a process called purifying selection,
and they do not contribute to the heterozygosity we observe. Loci that experience
purifying selection are said to be under selective constraint. The neutral mutation
rate for these loci (μn) is smaller than the total mutation rate (μ). In contrast, in many
noncoding regions, none of the mutations affect fitness, so in those regions the neu-
tral mutation rate is equal to the total mutation rate. Noncoding regions therefore
typically have higher heterozygosity, as predicted by
E q u a t i o n 7.1.
The same logic explains patterns of variation at
different sites within a coding locus. Many muta-
tions to the third positions of codons are syn-
onymous. Most mutations at the first and second
positions, however, are nonsynonymous. They
therefore have a lower neutral mutation rate than
do third positions, and as a result they are less vari-
able. The Adh locus in Figure 7.8 shows this effect:
only one of the 17 polymorphisms found in the
exons is nonsynonymous.
In sum, at sites of the genome that are free of
selection, all mutations are selectively neutral. These
mutations are free to drift through the population,
and they contribute to heterozygosity as they do.
But at sites that experience selection, many or most
mutations are deleterious. They are selected out of
the population and so contribute very little to hetero-
zygosity. These sites therefore tend to be less geneti-
cally variable.
Levels of polymorphism also vary systematically
along chromosomes. Regions with high recombi-
nation rates tend to be more polymorphic (FIGURE
7.9). We saw in Chapter 5 that a selective sweep of aFutuyma Kirkpatrick Evolution, 4e
Sinauer Associates
Troutt Visual Services
Evolution4e_07.09.ai Date 11-14-2016 01-18-17 01-24-17Heterozygosity (π)
TelomereTelomere Centromere TelomereChromosome 3Telomere CentromereChromosome 20.0000.0040.0080.0120.0000.0040.0080.012 4
3
2
1
0Recombination (cM/Mb)4
3
2
1
0FIGURE 7.9 Heterozygosity (π), shown by the points, along the left and
right arms of chromosomes 2 and 3 of Drosophila melanogaster. The
curves show the recombination rate (in centimorgans [cM] per Mb of
DNA) along the chromosome. Heterozygosity is reduced by background
selection and selective sweeps in regions of low recombination near the
centromeres and telomeres. (After [23].)07_EVOL4E_CH07.indd 175 3/23/17 9:09 AM