Transcript Models of molecular evolution III

Models of Molecular
Evolution III
Level 3 Molecular Evolution and
Bioinformatics
Jim Provan
Page and Holmes: Sections 7.5 – 7.8
The nearly neutral theory
Variable mutation rates
(generation time, metabolic rate, DNA repair)
Variable rates of
nucleotide substitution
Nearly neutral mutations
(population size and weak selection)
Development of the nearly neutral
theory
By the early 1970s it was
becoming clear that:
Some amino acid substitution
rates were inconsistent with
a Poisson clock
Levels of heterozygosity were
not as high as expected in
natural populations
Tomoko Ohta suggested
that most non-synonymous
changes were not perfectly
neutral
Deleterious
Advantageous
Neutral
Slightly deleterious
Slightly advantageous
Development of the nearly neutral
theory
Nearly neutral mutations are those where the product of
the population size and the selection coefficient is near
zero i.e. Ns  0
These changes are subject to weak natural selection as well as
genetic drift
Rate of substitution (per year) for nearly neutral mutations will
depend on population size, selective coefficient and mutation
rate, set to generation time
Mutations at non-coding and synonymous sites are still neutral
Relative importance of selection or genetic drift and
hence probability of fixation depends on population size
The nearly neutral theory and real
time molecular clocks
Long generation time
(lower mutation rate)
+
Short generation time
(higher mutation rate)
~ Real time
molecular clock
Small population size
(higher probability of fixation)
+
Large population size
(lower probability of fixation)
The nearly neutral theory and real
time molecular clocks
A study of 20 mammalian genes showed that R(t)
values were generally well over 1.0, rejecting the
Poisson clock
After correcting for generation time and lineage effects:
Average R(t) for synonymous sites dropped from 14.4 to 4.6
Average R(t) for non-synonymous sites only dropped from 8.26
to 6.95
Most variation in synonymous sites is due to lineage
effects
Non-synonymous rates are less generation time
dependent than synonymous sites
Testing the neutral theory within
species
Neutral theory makes two very important predictions
about levels of genetic variation within species:
Extent of polymorphism is a function only of the population
size (N) and the mutation rate (m)
Levels of polymorphism are correlated with amount of
variation between species i.e. genes that evolve slowly
between species also exhibit low variation within a species
Original allozyme studies cast doubt on these
predictions:
Levels of heterozygosity were found to be too low
Difficult to draw firm conclusions, since N and m are hard to
quantify, and allozymes underestimate diversity levels
Best to test theories at the DNA level
Testing the neutral theory within
species
Important assessment of the neutral theory is to test
proposed correlation between levels of within species
polymorphism and between species divergence:
If synonymous and non-synonymous substitutions are neutral
then ratio of both types of change will be the same within and
between species because they result from the same neutral
mutation process
Positive natural selection would alter this ratio because an
advantageous non-synonymous mutation would be fixed quicker
by natural selection i.e. be a polymorphism for less time, leading
to less within-species non-synonymous variation than expected
given levels detected between species
Patterns of substitution at the Adh
locus in Drosophila
D. melanogaster
2/14
D. simulans
0/11
1/2
D. yakuba
0/17
1/0
5/15
Synonymous
Non-synonymous
G = 7.43
Fixed Polymorphic
17
42
7
2
P = 0.006
Recombination and DNA
polymorphism in Drosophila
Most dramatic example of incompatibility between
levels of variation between and within a species
occurs in regions of the Drosophila genome where
recombination rates are low:
Distal tip of the X chromosome
Small chromosome IV
Neutralist explanation is that these regions have
either lower mutation rates or are under selective
constraint
If this is the case, these regions should also show
reduced levels of variation between species
Recombination and DNA
polymorphism in Drosophila
The yellow-achaete (y, ac) region on the X chromosome
of D. melanogaster has a reduced level of polymorphism
Extent of divergence in this region between D.
melanogaster and D. simulans (5.4%) is similar to that
observed in other genes (average 4.7%) – this
contradicts the neutral theory
Natural selection can explain this discrepancy through
the dual action of selective sweeps and genetic
hitchhiking
Hitchhiking and selective sweeps
t1
t2
t3
Time
t4
Can we resolve the neutralistselectionist debate?
Most support for neutral theory has come from
comparisons of genes across distantly related species
Natural selection is more apparent over shorter time scales
As time proceeds, fixation due to selection may be obscured by
neutral mutations
Majority of original selection events in adaptive radiations e.g.
mammals
Overall, it seems reasonable to conclude that both
selection and drift shape evolutionary fate of mutations:
Majority of substitutions do not affect fitness (neutralist)
The “footprints” of natural selection are still evident in more
recently evolved levels, particularly at non-synonymous sites