Transcript 3000_2013_2fg

you did about
this well
introducing the Bayesian posterior probability
of your exam grade!
more on mice
•
•
•
•
Darwin’s finches outstanding example
want to continue with this because we have
looked at 2 mice studies so far that have
inferred gene regions associated with color,
and have assumed selection
Darwin’s postulates: the trait is variable, the
trait is heritable...do we find out that survival
is variable, and associated with inheritance of
trait?
lets make sure it is what we think it is!
science isn’t always
complicated
recurring themes
•
•
•
•
how can a population better utilize available
resource? (some digestive enzyme)
how can individuals better camouflage
themselves? (fur or skin coloration)
how does an individual maximize probability of
offspring? (sperm motility)
you can start to generate hypotheses for what
could be ‘useful’ and then we recognize why
some patterns recur
•
•
•
phylogenetics and
biogeography
previous slide uses phylogeny of genes
unrelated to coat color to assess probability
that coat color changes happen
independently
biogeography: the study of how and why
organisms are distributed the way they are
phylogeny can help us see whether species
distributions represent descent from
common ancestral range or population
sticklebacks, redux
development of
pelvic spines in
marine populations
concomitant
reduction in
armoring in
freshwater
populations
genes responsible
• formation of lateral bony plates: QTL
mapping led to ecdysoplasin (Eda)
gene - involved with development of
adult integument and teeth in
vertebrates
• Eda
LOW homozygotes
have few lateral
plates vs. EdaHIGH homozygotes
• alleles different enough that they have
probably both existed for
are you a geneious?
•
•
•
informatics software makes it easier than
ever to explore what sequence data exist,
what they look like, how they are related to
one another
free version can do a lot; has tutorial for
basic exploration and bioinformatics
on wiki you might choose to explore allelic
variation, diversity of a population or
species, etc. (do so systematically, with
purpose...)
intron variation between
two allele types
(some heterozygotes)
high Eda alleles
low Eda alleles
“marine” types show up
repeatedly in freshwater
invasions; selection returns
population to less-armored
phenotype
agriculture
• human activities and behaviors have
led to domestication of animals for meat
and milk, crops for efficient nutrition
• know what the “green revolution” is?
• agriculture is a dramatic force of
selection
dairy consumption
(lactose tolerance,
lactase persistence)
more frequent in
groups with cultural
tendency to drink
milk
nature.com
signature of selection
• molecular evolution: analysis of how
DNA sequences evolve in response to
selection (and demographic) forces on
phenotypes (populations)
• allele frequencies and linkage
disequilibrium provide clues to the
history of a genome region
dna: site frequencies and
polymorphism
acctggctcgac
gtctggctcaac
acctagctcaat
acctagcccaac
acttggctcagc
acttggcttaac
acttggctcagc
acttggctcagc
•
how many segregating
sites?
•
how many are unique?
•
how many found in 2
individuals?
•
how many in 3, 4?
•
average # differences
between sequences
calculate ∏ and why
acctggctcgac
gtctggctcaac
acctagctcaat
acctagcccaac
acttggctcagc
acttggcttaac
acttggctcagc
acttggctcagc
•
eventually gets us to
estimating the population
size, with assumptions...
•
here, 3.03 is proportional
to the number of
individuals reproducing
times the mutation rate
(π=xNµ) thus (N=π/xµ)
how does that work?
•
briefly consider that there are N individuals in a
population
•
there is a rate µ at which a mutation arises at a locus
•
if N individuals reproduce successfully, we can expect
Nµ new mutations in next generation
•
diversity is scaled by the number of copies of a locus in
a reproductive event, e.g. there are 2 maternal and 2
paternal alleles that can have a mutation for a nuclear
diploid locus... x=4, so that π = 4Nµ
diversity at a locus
•
however, we often find that our estimates of population
size are far lower than the number we can count!!
failure of theory?
•
main cause of drift (coming after exam 1) is variance in
reproductive success (spatial, temporal, and individual
variance)
•
the effective population size accounts for changes in
population size, gender ratio, better nest sites, other
causes for some individuals contributing
disproportionately to next generation
Effective population
•
Several definitions, but
Ne=N/(variance) works,
where N is how many
individuals
•
If average individual has 1
offspring, most basic
assumption is variance of 1
(normal distribution), so
Ne=N
•
If variance high, Ne (and
diversity) goes down
can also partition the data
•
first codon position,
second position, third
position...
•
in protein-coding regions,
most variation is ‘silent’
or ‘synonymous’
•
why?
•
π is math equivalent to heterozygosity
proportion
of
• patterns to be explored in this diversity
individual
s
• more diversity in tropical species than polar?
heterozyg
• lower diversity following introduction?
ous at a
Things in Athens, Georgia
Procambarus clarkii, Town
Spring, π = 0.0
Notropis lutipinnis, Oconee
River, π = 2.1
signature of selection
(time passes)
width = length of homologous linked
segments around allele
we expect a classic signature
of “selective sweep” when
rare mutation quickly
increases
in frequency
would require strong
selection
one 2004 study
estimated up to 19% more
offspring surviving to maturity
Bersaglieri, T. et al. Am. J. Hum. Genet. 74, 1111–1120 (2004).
Nucleotide variation of studied regions on maize chromosome 10
Tian F et al. PNAS 2009;106:9979-9986
©2009 by National Academy of Sciences
blueridgeimpressions.wordpress.com
resistance - ability to prevent
damage (herbivore, herbicide,
toxin, pathogen)
tolerance - measure of fitness
following damage
you should be expecting this
answer by now!
•
•
•
•
pathways to
resistance
glyphosate attacks a highly conserved enzyme,
EPSPS, involved in amino acid construction
Monsanto developed transgenic crops that were
glyphosate-resistant; spray crops and only weeds
die?
1. some species have single amino acid difference in
EPSPS that render glyphosate useless
2. others have duplicated the EPSPS region and thus
produce more of this enzyme than glyphosate can
block!
•
•
•
•
•
•
interesting
interpretation
Bt is a toxin made by Bacillus thuringiensis; makes insects that eat
crops sick
spraying Bt on crops, or using transgenic crops that produce Bt,
reduces insect damage
but can insects evolve resistance? we would predict so
text (p244) notes that maintaining non-Bt refuges limits evolution of
resistance because resistance is costly (where there is no toxin,
resistant insects have lower fitness than non-resistant insects)
seems to be working; now it is law
why wouldn’t (or would) same approach work for antibiotic
resistance?
contemporary problem
39
•
atlantic cod - Gadus morhua
•
we’re fishing the last 0.01% of
this population...
•
fisheries are large uncontrolled experiments in evolution
•
rules put in place for what can be caught, where, what size, what species
•
rules can change, may be ignored
•
average 4-year-old cod is 10-15cm shorter now than 30 years ago: net size
selected against fast growth
•
heritability for rate of growth ~0.6
•
if reproductive capacity proportional to body size, population is now smaller
with a smaller potential rate of increase
40
bummer, man
fewer
fish...
smaller, younger
fish...
Hutchings &
Baum 2005
more on drift
what will happen?
•
serial colonization and founders effects
• new habitats?
• what do you predict?
49
change in allele frequency ∆p tells us
about the effective population size in
a different way from diversity (π)
the larger the population that is being
sampled from, the smaller the change
from one generation to next!
fixation
•
if allele frequency goes to 0 or 1, there is no
longer any variation, polymorphism,
segregation.... it STAYS there
• until mutation causes new polymorphism
• probability that any given allele goes to
fixation (frequency 1.0) is equal to its
frequency
– if
allele A is at frequency p=0.7, then there is 70%
chance that allele will fix
53
Starts at high
HETEROZYGOSITY
Ends with low
HETEROZYGOSITY
And 50/50 fixation of types
16
9!
effective population size
•
Ne (effective population size) is lower than N (actual
number of individuals) because of variance in
reproductive success
•
imbalance in males/females means not all of the more
common gender get to reproduce
•
“bottleneck” - a temporary reduction in population size
- means that all the diversity descends from a smaller
number of original individuals
•
different habitat quality, etc.
diversity is transient
•
without mutation, in a finite population
heterozygosity will decline every generation
– Hg+1
= Hg[1-(1/2N)]
– decline is faster with small N
•
selection can speed this loss (directional
selection) or slow it down (balancing,
frequency-dependent)
• migration, mating patterns also affect rate
58