Deviations from HWE

1
Deviations from HWE I. Mutation II.
Migration III. Non-Random Mating IV. Genetic
Drift V. The Neutral Theory
2
Deviations from HWE I. Mutation II.
Migration III. Non-Random Mating IV. Genetic
Drift V. The Neutral Theory A. Variation
1. Phenotypic variation was often interpreted as
having selective value in fact, most studies
confirmed that under one environmental condition
or another, there was a difference in fitness
among variations. Mayr (1963) "it is altogether
unlikely that two genes would have identical
selective value under all conditions under which
they may coexist in a population. Cases of
neutral polymorphism do not exist."
3
Deviations from HWE I. Mutation II.
Migration III. Non-Random Mating IV. Genetic
Drift V. The Neutral Theory A. Variation
1. Phenotypic variation was often interpreted as
having selective value in fact, most studies
confirmed that under one environmental condition
or another, there was a difference in fitness
among variations. Mayr (1963) "it is altogether
unlikely that two genes would have identical
selective value under all conditions under which
they may coexist in a population. Cases of
neutral polymorphism do not exist." 2. In the
1960's - lots of electrophoretic work revealed a
vast amount of variability - variability at the
gene or protein level that did not necessarily
correlate with morphological variation. Some are
silent mutations in DNA, or even neutral
substitution mutations. This variation results in
heterozygosity.
CCC Proline CCU Proline CCA Proline CCG
Proline
4
Deviations from HWE I. Mutation II.
Migration III. Non-Random Mating IV. Genetic
Drift V. The Neutral Theory A. Variation
3. Most populations showed mean
heterozygosities across ALL loci of about 10.
- And, about 20-30 of all loci are polymorphic
(have at least 2 alleles with frequencies over
1). Drosophila has 10,000 loci, so 3000 are
polymorphic. At these polymorphic loci, H .33
Conclusion - lots of variation at a genetic
level... is this also solely maintained by
selection?
5
Deviations from HWE I. Mutation II.
Migration III. Non-Random Mating IV. Genetic
Drift V. The Neutral Theory A. Variation B.
Genetic Load
6
Deviations from HWE I. Mutation II.
Migration III. Non-Random Mating IV. Genetic
Drift V. The Neutral Theory A. Variation B.
Genetic Load 1. "HARD" Selection can 'cost' a
population individuals
7
Deviations from HWE I. Mutation II.
Migration III. Non-Random Mating IV. Genetic
Drift V. The Neutral Theory A. Variation B.
Genetic Load 1. "HARD" Selection can 'cost' a
population individuals - those that die as a
consequence of differential fitness values.
8
Deviations from HWE I. Mutation II.
Migration III. Non-Random Mating IV. Genetic
Drift V. The Neutral Theory A. Variation B.
Genetic Load 1. "HARD" Selection can 'cost' a
population individuals - those that die as a
consequence of differential fitness values. -
the "breeding population" is smaller than the
initial population.
9
Deviations from HWE I. Mutation II.
Migration III. Non-Random Mating IV. Genetic
Drift V. The Neutral Theory A. Variation B.
Genetic Load 1. "HARD" Selection can 'cost' a
population individuals - those that die as a
consequence of differential fitness values. -
the "breeding population" is smaller than the
initial population. - Reproductive output must
compensate for this loss of individuals
10
Deviations from HWE I. Mutation II.
Migration III. Non-Random Mating IV. Genetic
Drift V. The Neutral Theory A. Variation B.
Genetic Load 1. "HARD" Selection can 'cost' a
population individuals - those that die as a
consequence of differential fitness values. -
the "breeding population" is smaller than the
initial population. - Reproductive output
must compensate for this loss of individuals -
The stronger the "hard" selection, the more
individuals are lost and the higher the
compensatory reproductive effort must be.
11
Deviations from HWE I. Mutation II.
Migration III. Non-Random Mating IV. Genetic
Drift V. The Neutral Theory A. Variation B.
Genetic Load 1. "HARD" Selection can 'cost' a
population individuals - those that die as a
consequence of differential fitness values. -
the "breeding population" is smaller than the
initial population. - - Reproductive output
must compensate for this loss of individuals -
The stronger the "hard" selection, the more
individuals are lost and the higher the
compensatory reproductive effort must be. - The
'cost' of replacing an allele with a new,
adaptive allele "Genetic Load" (L) L (optimal
fitness - mean fitness)/optimal fitness.
Essentially, this is a measure of the proportion
of individuals that will die as a consequence of
this "hard" selection. The lower the mean
fitness, the further the population is from the
optimum, and the more deaths there will be.
12

• B. Genetic Load
• 1. "HARD" Selection can 'cost' a population
  individuals
• 2. Why is this a problem?

13
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? - If variation is maintained by
selection, we are probably talking about
"heterosis" - selection for the heterozygote
where the heterozygote has the highest fitness
(and both alleles are maintained). (Selection
against the heterozygote can only maintain
variation at equilibrium, and this is
unstable).
14

• B. Genetic Load
• 1. "HARD" Selection can 'cost' a population
  individuals
• 2. Why is this a problem?
• - If variation is maintained by selection, we
  are probably talking about "heterosis" -
  selection for the heterozygote where the
  heterozygote has the highest fitness (and both
  alleles are maintained).
• - The problem is that load can be high in this
  situation, because lots of homozygotes are
  produced each generation, just to die by
  selection.

15

• B. Genetic Load
• 1. "HARD" Selection can 'cost' a population
  individuals
• 2. Why is this a problem?
• - If variation is maintained by selection, we
  are probably talking about "heterosis" -
  selection for the heterozygote where the
  heterozygote has the highest fitness (and both
  alleles are maintained).
• - The problem is that load can be high in this
  situation, because lots of homozygotes are
  produced each generation, just to die by
  selection.
• - Let's consider even a "best case" scenario

16

• B. Genetic Load
• 1. "HARD" Selection can 'cost' a population
  individuals
• 2. Why is this a problem?
• - If variation is maintained by selection, we
  are probably talking about "heterosis" -
  selection for the heterozygote where the
  heterozygote has the highest fitness (and both
  alleles are maintained).
• - The problem is that load can be high in this
  situation, because lots of homozygotes are
  produced each generation, just to die by
  selection.
• - Let's consider even a "best case" scenario
• - mean fitness 1 - H((st)/2)

17

• B. Genetic Load
• 1. "HARD" Selection can 'cost' a population
  individuals
• 2. Why is this a problem?
• - If variation is maintained by selection, we
  are probably talking about "heterosis" -
  selection for the heterozygote where the
  heterozygote has the highest fitness (and both
  alleles are maintained).
• - The problem is that load can be high in this
  situation, because lots of homozygotes are
  produced each generation, just to die by
  selection.
• - Let's consider even a "best case" scenario
• - mean fitness 1 - H((st)/2)
• - If s and t .1 (very weak), and H .33
  (average for Drosophila,
• above), then the mean fitness 0.967.

18
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? - If variation is maintained by
selection, we are probably talking about
"heterosis" - selection for the heterozygote
where the heterozygote has the highest fitness
(and both alleles are maintained). - The
problem is that load can be high in this
situation, because lots of homozygotes are
produced each generation, just to die by
selection. - Let's consider even a "best case"
scenario - mean fitness 1 - H((st)/2)
- If s and t .1 (very weak), and H .33
(average for Drosophila, above), then the
mean fitness 0.967. - Not bad not much
death due to selection in this situation...
19
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? - If variation is maintained by
selection, we are probably talking about
"heterosis" - selection for the heterozygote
where the heterozygote has the highest fitness
(and both alleles are maintained). - The
problem is that load can be high in this
situation, because lots of homozygotes are
produced each generation, just to die by
selection. - Let's consider even a "best case"
scenario - mean fitness 1 - H((st)/2)
- If s and t .1 (very weak), and H .33
(average for Drosophila, above), then the
mean fitness 0.967. - Not bad not much
death due to selection in this situation...
- HOWEVER, there are 3000 polymorphic loci
across the genome. So, mean fitness across the
genome (0.967)3000!
20
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? - If variation is maintained by
selection, we are probably talking about
"heterosis" - selection for the heterozygote
where the heterozygote has the highest fitness
(and both alleles are maintained). - The
problem is that load can be high in this
situation, because lots of homozygotes are
produced each generation, just to die by
selection. - Let's consider even a "best case"
scenario - mean fitness 1 - H((st)/2)
- If s and t .1 (very weak), and H .33
(average for Drosophila, above), then the
mean fitness 0.967. - Not bad not much
death due to selection in this situation...
- HOWEVER, there are 3000 polymorphic loci
across the genome. So, mean fitness across the
genome (0.967)3000! This becomes ridiculously
LOW (0.19 x 10-44) relative to the best case
genome that is heterozygous at every one of the
3000 loci. - So, some individuals die because
they are homozygous (and less fit) at A, others
die because they are homozygous (and less fit) at
B, other die because they are homozygous (and
less fit) at C, and so forth...
21
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? - If variation is maintained by
selection, we are probably talking about
"heterosis" - selection for the heterozygote
where the heterozygote has the highest fitness
(and both alleles are maintained). - The
problem is that load can be high in this
situation, because lots of homozygotes are
produced each generation, just to die by
selection. - Let's consider even a "best case"
scenario - mean fitness 1 - H((st)/2)
- If s and t .1 (very weak), and H .33
(average for Drosophila, above), then the
mean fitness 0.967. - Not bad not much
death due to selection in this situation... In
this case, the load is SO GREAT across the genome
that almost NOBODY lives to reproduce. And those
that do can not possibly produce enough offspring
to compensate for this amount of death.
22
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? So, hard selection can not be SOLELY
responsible for the variation we observe... a
population could not sustain itself under this
amount of genetic load...
23
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists -
Not all selection is "hard", imposing additional
deaths above background mortality.
24
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists -
Not all selection is "hard", imposing additional
deaths above background mortality. - There is
also "soft" selection, in which the death due to
selection occurs as a component of background
mortality, not in addition to it.
25
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists -
Not all selection is "hard", imposing additional
deaths above background mortality. - There is
also "soft" selection, in which the death due to
selection occurs as a component of background
mortality, not in addition to it. - For
instance, consider territoriality or competition
for a resource. Suppose there is only enough food
or space to support 50 individuals, but 60
offspring are produced each generation. Well,
each generation there are 10 deaths and there are
50 survivors".
26
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists -
Suppose we have a population of aa homozygotes
initially. All the territories are occupied by aa
individuals and 10 individuals die.
27
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists -
Suppose we have a population of aa homozygotes
initially. All the territories are occupied by aa
individuals and 10 individuals die. - Well, If
an 'A' allele is produce by mutation and
heterozygotes have the highest relative fitness
(probability of acquiring a territory), then the
allele "A" increase in frequency to
equilibrium....
28
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists -
Suppose we have a population of aa homozygotes
initially. All the territories are occupied by aa
individuals and 10 individuals die. - Well, If
an 'A' allele is produce by mutation and
heterozygotes have the highest relative fitness
(probability of acquiring a territory), then the
allele "A" increase in frequency to
equilibrium.... - Selection occurs, BUT THERE
ARE STILL ONLY 10 DEATHS PER GENERATION.
29
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists -
Suppose we have a population of aa homozygotes
initially. All the territories are occupied by aa
individuals and 10 individuals die. - Well, If
an 'A' allele is produce by mutation and
heterozygotes have the highest relative fitness
(probability of acquiring a territory), then the
allele "A" increase in frequency to
equilibrium.... - Selection occurs, BUT THERE
ARE STILL ONLY 10 DEATHS PER GENERATION. - In
this case there is NO genetic load, as selection
is NOT causing ADDITIONAL mortality. It is just
changing the probability of who dies.
30
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists -
Suppose we have a population of aa homozygotes
initially. All the territories are occupied by aa
individuals and 10 individuals die. - Well, If
an 'A' allele is produce by mutation and
heterozygotes have the highest relative fitness
(probability of acquiring a territory), then the
allele "A" increase in frequency to
equilibrium.... - Selection occurs, BUT THERE
ARE STILL ONLY 10 DEATHS PER GENERATION. - In
this case there is NO genetic load, as selection
is NOT causing ADDITIONAL mortality. It is just
changing the probability of who dies. - So,
selection across lots of loci does not
NECCESSARILY lead to impossible loads.... as long
as it is SOFT SELECTION
31
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists
b. Neutralists
32
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists
b. Neutralists - Maybe MOST of this variation
is NEUTRAL, and is simply maintained by drift as
new mutant alleles sequentially replace one
another.
33
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists
b. Neutralists - Maybe MOST of this variation
is NEUTRAL, and is simply maintained by drift as
new mutant alleles sequentially replace one
another. c. In a sense, the argument is
really about selection.
34
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists
b. Neutralists - Maybe MOST of this variation
is NEUTRAL, as is simply maintained by drift as
new mutant alleles sequentially replace one
another. c. In a sense, the argument is
really about selection. Selectionists state
that selection is important for 2 reasons - it
eliminates bad alleles and FAVORS advantageous
alleles.
35
B. Genetic Load 1. "HARD" Selection can 'cost' a
population individuals 2. Why is this a
problem? 3. Solutions a. Selectionists
b. Neutralists - Maybe MOST of this variation
is NEUTRAL, as is simply maintained by drift as
new mutant alleles sequentially replace one
another. c. In a sense, the argument is really
about selection. Selectionists state that
selection is important for 2 reasons - it
eliminates bad alleles and FAVORS advantageous
alleles. Neutralists agree that selection weeds
out deleterious alleles, but they claim that this
leaves a set of alleles that are functionally
equivalent - neutral - in relative value. And
changes in these equivalent alleles occur as a
consequence of drift.
36
Deviations from HWE I. Mutation II.
Migration III. Non-Random Mating IV. Genetic
Drift V. The Neutral Theory C. Neutral
Variation
Motoo Kimura 1924-1994
37
V. The Neutral Theory C. Neutral Variation -
Variation occurs at many levels, from genes to
proteins to physical and behavioral
characteristics of organisms.
38
V. The Neutral Theory C. Neutral Variation -
Variation occurs at many levels, from genes to
proteins to physical and behavioral
characteristics of organisms. - adaptive
phenotypic variation is due to selection.
39
V. The Neutral Theory C. Neutral Variation -
Variation occurs at many levels, from genes to
proteins to physical and behavioral
characteristics of organisms. - adaptive
phenotypic variation is due to selection. -
But is ALL genetic variation of selective value?
40
V. The Neutral Theory C. Neutral Variation -
Variation occurs at many levels, from genes to
proteins to physical and behavioral
characteristics of organisms. - adaptive
phenotypic variation is due to selection. -
But is ALL genetic variation of selective value?
- "no" obviously, silent mutations are not
maintained by selection
41
V. The Neutral Theory C. Neutral Variation -
Variation occurs at many levels, from genes to
proteins to physical and behavioral
characteristics of organisms. - adaptive
phenotypic variation is due to selection. -
But is ALL genetic variation of selective value?
- "no" obviously, silent mutations are not
maintained by selection - So, Kimura suggested
that there is too much variation at the DNA level
to be explained by selection... he suggested that
MOST of the variation in DNA is of NO selective
value - it is NEUTRAL VARIATION.
42
V. The Neutral Theory C. Neutral Variation -
Variation occurs at many levels, from genes to
proteins to physical and behavioral
characteristics of organisms. - adaptive
phenotypic variation is due to selection. -
But is ALL genetic variation of selective value?
- "no" obviously, silent mutations are not
maintained by selection - So, Kimura suggested
that there is too much variation at the DNA level
to be explained by selection... he suggested that
MOST of the variation in DNA is of NO selective
value - it is NEUTRAL VARIATION. - Curiously,
the rate of replacement by drift, alone the
rate of mutation
43
V. The Neutral Theory C. Neutral Variation -
Variation occurs at many levels, from genes to
proteins to physical and behavioral
characteristics of organisms. - adaptive
phenotypic variation is due to selection. -
But is ALL genetic variation of selective value?
- "no" obviously, silent mutations are not
maintained by selection - So, Kimura suggested
that there is too much variation at the DNA level
to be explained by selection... he suggested that
MOST of the variation in DNA is of NO selective
value - it is NEUTRAL VARIATION. - Curiously,
the rate of replacement by drift, alone the
rate of mutation 1) The number of new alleles
produced at a locus 2N(m), where m is the
mutation rate.
44
V. The Neutral Theory C. Neutral Variation -
Variation occurs at many levels, from genes to
proteins to physical and behavioral
characteristics of organisms. - adaptive
phenotypic variation is due to selection. -
But is ALL genetic variation of selective value?
- "no" obviously, silent mutations are not
maintained by selection - So, Kimura suggested
that there is too much variation at the DNA level
to be explained by selection... he suggested that
MOST of the variation in DNA is of NO selective
value - it is NEUTRAL VARIATION. - Curiously,
the rate of replacement by drift, alone the
rate of mutation 1) The number of new alleles
produced at a locus 2N(m), where m is the
mutation rate. - So, if the average mutation
rate is 1 in 10,000, but there are 20,000
individuals (2N 40,000 alleles), then on
average 4 new alleles will be produced by
mutation every generation.
45
V. The Neutral Theory C. Neutral Variation -
Variation occurs at many levels, from genes to
proteins to physical and behavioral
characteristics of organisms. - adaptive
phenotypic variation is due to selection. -
But is ALL genetic variation of selective value?
- "no" obviously, silent mutations are not
maintained by selection - So, Kimura suggested
that there is too much variation at the DNA level
to be explained by selection... he suggested that
MOST of the variation in DNA is of NO selective
value - it is NEUTRAL VARIATION. - Curiously,
the rate of replacement by drift, alone the
rate of mutation 1) The number of new alleles
produced at a locus 2N(m), where m is the
mutation rate. - So, if the average mutation
rate is 1 in 10,000, but there are 20,000
individuals (2N 40,000 alleles), then on
average 4 new alleles will be produced by
mutation every generation. 2) Each allele has a
probability of fixation 1/2N.
46
V. The Neutral Theory C. Neutral Variation -
Variation occurs at many levels, from genes to
proteins to physical and behavioral
characteristics of organisms. - adaptive
phenotypic variation is due to selection. -
But is ALL genetic variation of selective value?
- "no" obviously, silent mutations are not
maintained by selection - So, Kimura suggested
that there is too much variation at the DNA level
to be explained by selection... he suggested that
MOST of the variation in DNA is of NO selective
value - it is NEUTRAL VARIATION. - Curiously,
the rate of replacement by drift, alone the
rate of mutation 1) The number of new alleles
produced at a locus 2N(m), where m is the
mutation rate. - So, if the average mutation
rate is 1 in 10,000, but there are 20,000
individuals (2N 40,000 alleles), then on
average 4 new alleles will be produced by
mutation every generation. 2) Each allele has a
probability of fixation 1/2N. 3) So, the rate
of replacement (number of new alleles formed) x
(probability one become fixed) 2N(m) x 1/2N m
per generation.
47
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results
48
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions
49
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions - Rates should vary in
different codon positions. Variation at the third
position should be higher, because these are
usually silent mutations. Mutations at the first
two position change amino acids, and these
changes are deleterious.
50
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions - Rates should vary in
different codon positions. Variation at the third
position should be higher, because these are
usually silent mutations. Mutations at the
second position change amino acids, and these
changes are deleterious. PATTERN CONFIRMED.
51
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions - Rates should vary in
different codon positions. Variation at the third
position should be higher, because these are
usually silent mutations. Mutations at the
second position change amino acids, and these
changes are deleterious. PATTERN CONFIRMED. -
Rates should vary in coding and non-coding
regions. Variation in Introns should occur more
rapidly than variation in exons, since introns
are not transcribed and are also invisible to
selection.
52
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions - Rates should vary in
different codon positions. Variation at the third
position should be higher, because these are
usually silent mutations. Mutations at the
second position change amino acids, and these
changes are deleterious. PATTERN CONFIRMED. -
Rates should vary in coding and non-coding
regions. Variation in Introns should occur more
rapidly than variation in exons, since introns
are not transcribed and are also invisible to
selection. PATTERN CONFIRMED
53
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions - Rates should vary in
different codon positions. Variation at the third
position should be higher, because these are
usually silent mutations. Mutations at the
second position change amino acids, and these
changes are deleterious. PATTERN CONFIRMED. -
Rates should vary in coding and non-coding
regions. Variation in Introns should occur more
rapidly than variation in exons, since introns
are not transcribed and are also invisible to
selection. PATTERN CONFIRMED - Rates should
vary in functional and non-functional regions of
proteins.
54
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions - Rates should vary in
different codon positions. Variation at the third
position should be higher, because these are
usually silent mutations. Mutations at the
second position change amino acids, and these
changes are deleterious. PATTERN CONFIRMED. -
Rates should vary in coding and non-coding
regions. Variation in Introns should occur more
rapidly than variation in exons, since introns
are not transcribed and are also invisible to
selection. PATTERN CONFIRMED - Rates should
vary in functional and non-functional regions of
proteins. PATTERN CONFIRMED
55
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions - Rates should vary in
different codon positions. Variation at the third
position should be higher, because these are
usually silent mutations. Mutations at the
second position change amino acids, and these
changes are deleterious. PATTERN CONFIRMED. -
Rates should vary in coding and non-coding
regions. Variation in Introns should occur more
rapidly than variation in exons, since introns
are not transcribed and are also invisible to
selection. PATTERN CONFIRMED - Rates should
vary in functional and non-functional regions of
proteins. PATTERN CONFIRMED - Rates should
vary between vital proteins and less vital
proteins.
56
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions - Rates should vary in
different codon positions. Variation at the third
position should be higher, because these are
usually silent mutations. Mutations at the
second position change amino acids, and these
changes are deleterious. PATTERN CONFIRMED. -
Rates should vary in coding and non-coding
regions. Variation in Introns should occur more
rapidly than variation in exons, since introns
are not transcribed and are also invisible to
selection. PATTERN CONFIRMED - Rates should
vary in functional and non-functional regions of
proteins. PATTERN CONFIRMED - Rates should
vary between vital proteins and less vital
proteins. PATTERN CONFIRMED
57
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions 2. Rates of replacement
(substitution of one fixed allele by another that
reaches fixation) should be constant over
geologic time.
58

• V. The Neutral Theory
• C. Neutral Variation
• D. Predictions and Results
• 1. Rates of molecular evolution should vary in
  functional and non-functional regions
• 2. Rates of replacement (substitution of one
  fixed allele by another that reaches fixation)
  should be constant over geologic time.
• - If changes are random and occur at a given
  rate, then they should "tick" along like a clock.
  

59
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions 2. Rates of replacement
(substitution of one fixed allele by another that
reaches fixation) should be constant over
geologic time. - If changes are random and
occur at a given rate, then they should "tick"
along like a clock. - Selection should slow
change when an adapted complex occurs, and speed
rates when a new adaptive combination occurs,
like in obviously adaptive morphological traits.
- PATTERNS CONFIRMED (usually).
60
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions 2. Rates of replacement
(substitution of one fixed allele by another that
reaches fixation) should be constant over
geologic time. - If changes are random and
occur at a given rate, then they should "tick"
along like a clock. - Selection should slow
change when an adapted complex occurs, and speed
rates when a new adaptive combination occurs,
like in obviously adaptive morphological traits.
- PATTERNS CONFIRMED (usually).
61

• V. The Neutral Theory
• C. Neutral Variation
• D. Predictions and Results
• 1. Rates of molecular evolution should vary in
  functional and non-functional regions
• 2. Rates of replacement (substitution of one
  fixed allele by another that reaches fixation)
  should be constant over geologic time.
• 3. Rates of morphological change should be
  independent of the rate of molecular change.

62
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results 1. Rates of molecular
evolution should vary in functional and
non-functional regions 2. Rates of replacement
(substitution of one fixed allele by another that
reaches fixation) should be constant over
geologic time. 3. Rates of morphological change
should be independent of the rate of molecular
change. - "Living Fossils" show extreme
genetic change and variation, yet have remained
morphologically unchanged for millenia. And, the
rate of genetic change in this morphologically
constant species is the same as in hominids,
which have changed dramatically in morphology
over a short period.
63

• V. The Neutral Theory
• C. Neutral Variation
• D. Predictions and Results
• 1. Rates of molecular evolution should vary in
  functional and non-functional regions
• 2. Rates of replacement (substitution of one
  fixed allele by another that reaches fixation)
  should be constant over geologic time.
• 3. Rates of morphological change should be
  independent of the rate of molecular change.
• - "Living Fossils" show extreme genetic change
  and variation, yet have remained morphologically
  unchanged for millenia. And, the rate of genetic
  change in this morphologically constant species
  is the same as in hominids, which have changed
  dramatically in morphology over a short period.
  PATTERN CONFIRMED

64
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions
65
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions 1. Selection also explains different
mutation rates in functional and non-functional
regions
66
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions 1. Selection also explains different
mutation rates in functional and non-functional
regions Essentially, since most adaptive changes
should be slight, fewer mutations in functional
regions are likely to improve function. So, the
rate of change is "constrained" to only those
changes that are neutral or ADAPTIVE. Also, a
change of one AA is likely to cause a smaller
change if it is in a less functional region.
"Tweeking" less functional regions might be
adaptive, whereas "tweeking" functional regions
are more likely to be deleterious.
67
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions 1. Selection also explains different
mutation rates in functional and non-functional
regions 2. A truly neutral clock should tick
off mutations at a constant rate. But should this
ticking occur per unit time, or per generation?
68
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions 1. Selection also explains different
mutation rates in functional and non-functional
regions 2. A truly neutral clock should tick
off mutations at a constant rate. But should this
ticking occur per unit time, or per generation?
- Since mutations produce new alleles (a new
"tick"), and mutations only occur during
replication of the DNA, it would seem that a
truly neutral clock should tick at a rate
dependent on the generation time of the organism.
69
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions 1. Selection also explains different
mutation rates in functional and non-functional
regions 2. A truly neutral clock should tick
off mutations at a constant rate. But should this
ticking occur per unit time, or per generation?
- Since mutations produce new alleles (a new
"tick"), and mutations only occur during
replication of the DNA, it would seem that a
truly neutral clock should tick at a rate
dependent on the generation time of the
organism. - Species with rapid generation times
should accumulate mutations at a faster rate than
long-lived species with slower generation times.
70
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions 1. Selection also explains different
mutation rates in functional and non-functional
regions 2. A truly neutral clock should tick
off mutations at a constant rate. But should this
ticking occur per unit time, or per generation?
- Since mutations produce new alleles (a new
"tick"), and mutations only occur during
replication of the DNA, it would seem that a
truly neutral clock should tick at a rate
dependent on the generation time of the
organism. - Species with rapid generation times
should accumulate mutations at a faster rate than
long-lived species with slower generation
times. - This is true of non-coding DNA... but
not true for proteins, as we have seen. Proteins
accumulate mutations in absolute time, not
generational time.
71
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions 1. Selection also explains different
mutation rates in functional and non-functional
regions 2. A truly neutral clock should tick
off mutations at a constant rate. But should this
ticking occur per unit time, or per generation?
- Since mutations produce new alleles (a new
"tick"), and mutations only occur during
replication of the DNA, it would seem that a
truly neutral clock should tick at a rate
dependent on the generation time of the
organism. - Species with rapid generation times
should accumulate mutations at a faster rate than
long-lived species with slower generation
times. - This is true of non-coding DNA... but
not true for proteins, as we have seen. Proteins
accumulate mutations in absolute time, not
generational time. THIS IS INCONSISTENT WITH THE
NEUTRAL MODEL
72
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions F. The Nearly Neutral Model
(Ohta)
73
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions F. The Nearly Neutral Model
(Ohta) - Ohta included the very weak effect
against slightly deleterious mutations. He found
that, if s lt 1/2Ne, then alleles are essentially
neutral and become fixed as drift would predict.
74
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions F. The Nearly Neutral Model
(Ohta) - Ohta included the very weak effect
against slightly deleterious mutations. He found
that, if s lt 1/2Ne, then alleles are essentially
neutral and become fixed as drift would predict.
- In other words, in small populations, drift
predominates unless selection is fairly strong
(in a population of Ne 5, drift will
predominante unless s gt 0.1).
75
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions F. The Nearly Neutral Model
(Ohta) - Ohta included the very weak effect
against slightly deleterious mutations. He found
that, if s lt 1/2Ne, then alleles are essentially
neutral and become fixed as drift would predict.
- In other words, in small populations, drift
predominates unless selection is fairly strong
(in a population of Ne 5, drift will
predominante unless s gt 0.1). - In large
populations, selection predominates, even if it
is fairly weak (if Ne 10,000, then selection
will predominate if s gt 0.00005).
76
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions F. The Nearly Neutral Model
(Ohta) SO..... (GET READY FOR THIS!!!!)
77
F. The Nearly Neutral Model (Ohta) SO. - We
observe a constant AA substitution rate across
species, even though we would expect that species
with shorter generation times should have FASTER
rates of substitution.
OBS.
Sub. Rate
EXP.
Short GEN TIME Long
78
F. The Nearly Neutral Model (Ohta) SO. - We
observe a constant AA substitution rate across
species, even though we would expect that species
with shorter generation times should have FASTER
rates of substitution. - So, something must be
'slowing down' this rate of substitution in
species with short gen. times. What's slowing it
down is their large populations size, such that
the effects of drift, alone, are reduced.
LARGE POP. SIZE
OBS.
Sub. Rate
EXP.
Short GEN TIME Long
79
F. The Nearly Neutral Model (Ohta) SO. - We
observe a constant AA substitution rate across
species, even though we would expect that species
with shorter generation times should have FASTER
rates of substitution. - So, something must be
'slowing down' this rate of substitution in
species with short gen. times. What's slowing it
down is their large populations size, such that
the effects of drift, alone, are reduced. -
Likewise, species with long generation times have
small populations, and substitution by drift and
fixation is more rapid than expected based on
generation time, alone.
SMALL POP. SIZE
OBS.
Sub. Rate
EXP.
Short GEN TIME Long
80
F. The Nearly Neutral Model (Ohta) SO. - We
observe a constant AA substitution rate across
species, even though we would expect that species
with shorter generation times should have FASTER
rates of substitution. - So, something must be
'slowing down' this rate of substitution in
species with short gen. times. What's slowing it
down is their large populations size, such that
the effects of drift, alone, are reduced. -
Likewise, species with long generation times have
small populations, and substitution by drift and
fixation is more rapid than expected based on
generation time, alone.
SMALL POP. SIZE
SO. - The constant rate of AA substitution
across species is due to the balance between
generation time and population size.
OBS.
Sub. Rate
EXP.
Short GEN TIME Long
81
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions F. The Nearly Neutral Model
(Ohta) SO. - The constant rate of AA
substitution across species with different
generation times is due to the counter-balancing
effect of population size, which is inversely
correlated with generation time.
82
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions F. The Nearly Neutral Model (Ohta)
G. Conclusions
83
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions F. The Nearly Neutral Model (Ohta)
G. Conclusions - Neutral variability
certainly exists in non-coding DNA, especially.

84
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions F. The Nearly Neutral Model (Ohta)
G. Conclusions - Neutral variability
certainly exists in non-coding DNA, especially.
- However, it is possible that selection
maintains molecular variation as well,
particularly in coding regions.
85
V. The Neutral Theory C. Neutral Variation
D. Predictions and Results E. Problems and
Resolutions F. The Nearly Neutral Model (Ohta)
G. Conclusions - Neutral variability
certainly exists in non-coding DNA, especially.
- However, it is possible that selection
maintains molecular variation as well,
particularly in coding regions. It is also
possible that selection maintains variability in
non-coding regions, as well, if these are
"functional" in a structural or regulatory
manner.
86
IV. Selection and Other Factors A.
Mutation
87
IV. Selection and Other Factors A. Mutation
- Mutation can maintain a deleterious allele in
the population against the effects of selection,
such that
88
IV. Selection and Other Factors A. Mutation
- Mutation can maintain a deleterious allele in
the population against the effects of selection,
such that q(eq) v(m/s)
89
IV. Selection and Other Factors A. Mutation
- Mutation can maintain a deleterious allele in
the population against the effects of selection,
such that q(eq) v(m/s) - more deleterious
alleles are maintained if m increases, or if
selection differential declines... (if it is not
that bad to have it)
90
IV. Selection and Other Factors A. Mutation
B. Drift and "Adaptive Landscapes" 1. Single
Locus - consider a locus with selection
against the heterozygote
p 0.4, q 0.6 AA Aa aa
Parental "zygotes" 0.16 0.48 0.36 1.00
prob. of survival (fitness) 0.8 0.4 0.6
Relative Fitness 1 0.5 0.75
Corrected Fitness 1 0.5 1.0 1 0.25
formulae 1 s 1 t
peq t/(s t) .25/.75 0.33
91
IV. Selection and Other Factors A. Mutation
B. Drift and "Adaptive Landscapes" 1. Single
Locus - consider a locus with selection
against the heterozygote
1.0
0.75
mean fitness
1.0
0
0.33
92
IV. Selection and Other Factors A. Mutation
B. Drift and "Adaptive Landscapes" 1. Single
Locus - suppose there is random movement up
the 'wrong' slope?
1.0
0.75
mean fitness
1.0
0
0.33
93
IV. Selection and Other Factors A. Mutation
B. Drift and "Adaptive Landscapes" 1. Single
Locus
1.0
- if this is a large pop with no drift, the
population will become fixed on the 'suboptimal'
peak, (p 0, q 1.0, w 0.75).
0.75
mean fitness
1.0
0
0.33
94
IV. Selection and Other Factors A. Mutation
B. Drift and "Adaptive Landscapes" 1. Single
Locus
1.0
- BUT if it is small, then drift may be
important... because only DRIFT can randomly
BOUNCE the gene freq's to the other slope!
0.75
mean fitness
1.0
0
0.33
95
IV. Selection and Other Factors A. Mutation
B. Drift and "Adaptive Landscapes" 1. Single
Locus
1.0
- And then selection can push the pop up the most
adaptive slope!
0.75
mean fitness
1.0
0
0.33
96
IV. Selection and Other Factors A. Mutation
B. Drift and "Adaptive Landscapes" 1. Single
Locus
1.0
- The more shallow the 'maladaptive valley',
(representing weaker selection differentials) the
easier it is for drift to cross it...
0.75
mean fitness
1.0
0
0.33
97
IV. Selection and Other Factors A. Mutation
B. Drift and "Adaptive Landscapes" 2. Two Loci
- create a 3-D landscape, with "mean fitness" as
the 'topographic relief"
Suppose AAbb and aaBB work well, but combinations
of the two do not (epistatic, like butterfly
mimicry).
1.0
f(A)
1.0
f(B)
98
IV. Selection and Other Factors A. Mutation
B. Drift and "Adaptive Landscapes" 2. Two Loci
- create a 3-D landscape, with "mean fitness" as
the 'topographic relief"
Again, strong selection or weak drift will cause
mean fitness to move up nearest slopes.
1.0
f(A)
1.0
f(B)
99
IV. Selection and Other Factors A. Mutation
B. Drift and "Adaptive Landscapes" 2. Two Loci
- create a 3-D landscape, with "mean fitness" as
the 'topographic relief"
Only strong drift or weak selection and some
drift (shallow valley) can cause the population
to cross the maladaptive valley.
1.0
f(A)
1.0
f(B)
100
IV. Selection and Other Factors A. Mutation
B. Drift and "Adaptive Landscapes" - so, the
interactions between drift and selection are
necessary for a population to find the optimal
adaptive peak... think about this in the context
of peripatric speciation.... THINK HARD about
this...