gENETIC dRIfT: EvolUTIoN AT RANdoM 183
accumulate differences more rapidly than will sites that experi-
ence strong purifying selection. A natural experiment to test
this idea comes from pseudogenes, which are nonfunctional
duplicates of functioning genes. Because pseudogenes do not
produce a gene product, we expect that mutations anywhere in
their DNA sequences will be selectively neutral, and they should
therefore evolve rapidly. Consistent with Kimura’s neutral theory,
pseudogenes are among the fastest evolving parts of the genome
(FIGURE 7.19). The neutral theory also predicts that inside coding
regions, synonymous differences should accumulate faster than
nonsynonymous differences among species. Again, that is what
the data show (see Figure 7.18).
The neutral theory does a good job of explaining why parts of
the genome that experience little or no purifying selection evolve
so fast. We now know, however, that adaptive evolution is much
more important to molecular evolution than Kimura believed. We
saw earlier that in some organisms, fully half of the amino acid differences between
closely related species result from positive selection rather than drift [8, 16].
If adaptive evolution is so common, how can we explain molecular clocks?
Fluctuating selection pressures, for example caused by changing environments,
can also cause a locus to evolve in a relatively steady, clocklike way. A relatively
constant appearance of new beneficial mutations in a population will produce the
same result. Clocklike evolution has been seen in experimental populations of E.
coli, where it was possible to show that the genetic changes resulted from positive
selection and not drift [3]. Some genes evolve at very uneven rates and so make
poor molecular clocks. But those genes that do evolve at relatively constant rates,
whether by selection or drift, give us valuable tools to estimate divergence times
among species and to build phylogenies.
In sum, the differences among the DNA sequences of species evolve by both drift
and by positive selection. Noncoding regions of the genome that are free of purify-
ing selection accumulate differences quickly, as predicted by the neutral theory of
molecular evolution. In coding regions, many sites are under purifying selection. Like
neutral sites, they accumulate differences by drift, but more slowly. But other sites in
coding regions evolve largely by positive selection, not drift. We will now look at how
we can determine which changes occurred by adaptation and which by drift.Searching the genes for Signatures
of AdaptationOur genes are books that have recorded the evolutionary histories of our ances-
tors. Human geneticists have recently made great progress in decoding these
books. One important goal is to find genes that show signs of recent adaptive
evolution. These will help us understand the genetic basis of adaptation, how we
became different from other species, and how differences among human popula-
tions evolved.
The biggest challenge in finding signs of recent positive selection in DNA
sequences is that other evolutionary processes are also operating. Because genetic
drift is always at work, drift provides a null model. If we see that the pattern of
polymorphism within a species, or differences among species, differs from what
we expect under drift alone, that suggests selection may have played a role. Evolu-
tionary biologists have developed a toolbox of methods to distinguish the signals
of selection and drift. We discussed one of the approaches in Chapter 5, where weFutuyma Kirkpatrick Evolution, 4e
Sinauer Associates
Troutt Visual Services
Evolution4e_07.19.ai Date 11-11-2016 01-18-17PseudogenesIntergenic regionsIntronsCoding0 5 10 15
Evolutionary rate20FIGURE 7.19 The relative rates of evolution for different
kinds of DNA, estimated from differences between humans
and chimpanzees. These rates are measured as the number of
differences per nucleotide site, multiplied by 10^3. (After [11].)07_EVOL4E_CH07.indd 183 3/23/17 9:09 AM