Lecture 7. Neutral evolution, the problem of polymorphism and
the strength of selection
Following the rediscover of Mendel's work and following numerous
examples where selection could be seen to be operating (the rise of the
melanic form of the peppered moth being but one), there developed a new
understanding of natural selection and its operation. This new
understanding is known as the "modern synthesis". As the precise
formulation of the idea of natural selection isn't quite the same as
Darwin's original formulation, the new fusion of genetics, maths and
Darwin's idea of selection is known as neo-Darwinism. The main
developers of neo-Darwinism were three mathematical geneticists: RA
Fisher, JBS Haldane and Sewall Wright.
In the 1960s-70s, the maths of selection was challenged by some
unexpected findings:
1. Polymorphism is common: people (notably Dick Lewontin) in 1950s
and 1960 started looking at polymorphisms. Lewontin ad Hubby first
applied acrylamide gel electrophoresis and fond that 39% of D.
pseudobscura’s proteins are polymorphic. A polymorphism is the
existence of more than one allele at a locus. Using sequencing
methods we can now detect these differences directly. Previously,
people could only examine the polymorphism by examining the
proteins. Protein electrophoresis involves the migration of proteins
under the influence of an electric field. Different versions of the same
locus (i.e. different allelic forms of the protein) can be identified by this
means. These different forms are known as allozymes. Procedures
such as starch gel protein electrophoresis therefore allowed the first
glimpse at the levels of polymorphism seen in natural populations.
A very low level of polymorphism is expected when mutation and
selection are balanced. However investigations of allozymes suggested
that the different variants could be very common. The calculations of
the mathematical geneticists suggested that polymorphism like this
should not be especially common - in simple one locus population
genetics one expects polymorphism only under special conditions:
overdominance, frequency dependent selection etc. There was then a
discrepancy between the simple selection models and the apparent
amount of polymorphism. How could the circle be squared?
2. Genes appear to be evolving too fast:
, With the amino acid sequence of alpha and beta haemoglobin,
cytochrome c and triosephosphate isomerase, from humans and other
mammals Kimura (1968) estimated that the average 100 amino acid
protein sees one substitution every 28 x 104 yrs.
Multiplying this up across the genome, he estimates that there must be
one nucleotide substitution every two years.
But Haldane had noted that as selection requires mortality, there will
be a limit to the amount of selectively driven change - a maximum of
one every 300 years.
Kimura thus argued that the total amount of change couldn’t be
owing to selection as there would need to be too much selective
death.
Kimura's theory of neutral evolution
What if our two variants have no effects on fitness? Neither will
deterministically spread. But we can still get changing allele frequencies
owing to chance.
In the late sixties the Japanese mathematical Geneticist Motoo Kimura
suggested an answer to this dilemma of the excess of polymorphism. He
argued that for many variants at the molecular level selection doesn't
differentiate between different types. Why then should they become
common – the answer Kimura said is simple – chance alone. As the two
variants that we are to consider are not associated with a selective
difference they are known as neutral alleles and the process of their
evolution is neutral evolution.
Kimura's insight was that that majority of the maths laid about under neo-
Darwinism looked at the average change in frequency on an allele. So,
for example, on the average a selectively favourable allele increases in
frequency. But Kimura argued that the variance is also important.
Consider for example two alleles A and a that selection considers equal.
Imagine 1000 populations starting off with 50:50 ratios of these two. On
the average next generation there will still be a 50:50 ratio. However,
the strength of selection
Following the rediscover of Mendel's work and following numerous
examples where selection could be seen to be operating (the rise of the
melanic form of the peppered moth being but one), there developed a new
understanding of natural selection and its operation. This new
understanding is known as the "modern synthesis". As the precise
formulation of the idea of natural selection isn't quite the same as
Darwin's original formulation, the new fusion of genetics, maths and
Darwin's idea of selection is known as neo-Darwinism. The main
developers of neo-Darwinism were three mathematical geneticists: RA
Fisher, JBS Haldane and Sewall Wright.
In the 1960s-70s, the maths of selection was challenged by some
unexpected findings:
1. Polymorphism is common: people (notably Dick Lewontin) in 1950s
and 1960 started looking at polymorphisms. Lewontin ad Hubby first
applied acrylamide gel electrophoresis and fond that 39% of D.
pseudobscura’s proteins are polymorphic. A polymorphism is the
existence of more than one allele at a locus. Using sequencing
methods we can now detect these differences directly. Previously,
people could only examine the polymorphism by examining the
proteins. Protein electrophoresis involves the migration of proteins
under the influence of an electric field. Different versions of the same
locus (i.e. different allelic forms of the protein) can be identified by this
means. These different forms are known as allozymes. Procedures
such as starch gel protein electrophoresis therefore allowed the first
glimpse at the levels of polymorphism seen in natural populations.
A very low level of polymorphism is expected when mutation and
selection are balanced. However investigations of allozymes suggested
that the different variants could be very common. The calculations of
the mathematical geneticists suggested that polymorphism like this
should not be especially common - in simple one locus population
genetics one expects polymorphism only under special conditions:
overdominance, frequency dependent selection etc. There was then a
discrepancy between the simple selection models and the apparent
amount of polymorphism. How could the circle be squared?
2. Genes appear to be evolving too fast:
, With the amino acid sequence of alpha and beta haemoglobin,
cytochrome c and triosephosphate isomerase, from humans and other
mammals Kimura (1968) estimated that the average 100 amino acid
protein sees one substitution every 28 x 104 yrs.
Multiplying this up across the genome, he estimates that there must be
one nucleotide substitution every two years.
But Haldane had noted that as selection requires mortality, there will
be a limit to the amount of selectively driven change - a maximum of
one every 300 years.
Kimura thus argued that the total amount of change couldn’t be
owing to selection as there would need to be too much selective
death.
Kimura's theory of neutral evolution
What if our two variants have no effects on fitness? Neither will
deterministically spread. But we can still get changing allele frequencies
owing to chance.
In the late sixties the Japanese mathematical Geneticist Motoo Kimura
suggested an answer to this dilemma of the excess of polymorphism. He
argued that for many variants at the molecular level selection doesn't
differentiate between different types. Why then should they become
common – the answer Kimura said is simple – chance alone. As the two
variants that we are to consider are not associated with a selective
difference they are known as neutral alleles and the process of their
evolution is neutral evolution.
Kimura's insight was that that majority of the maths laid about under neo-
Darwinism looked at the average change in frequency on an allele. So,
for example, on the average a selectively favourable allele increases in
frequency. But Kimura argued that the variance is also important.
Consider for example two alleles A and a that selection considers equal.
Imagine 1000 populations starting off with 50:50 ratios of these two. On
the average next generation there will still be a 50:50 ratio. However,
