populations
DESCRIPTION
Musings on what is a populationTRANSCRIPT
What is a Population?
Bob O'HaraDepartment of Mathematics and Statistics
University of HelsinkiFinland
The population is a basic concept in ecology and evolutionary biology
The definition of a population is not precise
Recently, methods have been developed which have been used to
infer the existence of populations
How useful are they?
An Old Observation
Normally the ‘‘population’’ is more or less an abstraction
because there is a considerable interchange of individuals between neighboring populations,
owing to the absence or incompleteness of physical barriers.
Mayr, 1942
Lisa Gannett suggests that populations are not “real”
Populations are pragmatically and variably constituted in different sorts of investigations of species genome diversity.
Population boundaries are not fixed but vary fromone context of inquiry to another.
Gannett, 2003
If populations are real, then all methods for defining them should
give us similar outcomes
If different methods give different answers, this suggests that populations are not “mind
independent”
A difference in definitions
Local, partially isolated breeding groupsMaynard Smith, 1998
The group is primary
Individuals are primary
a group of organisms of the same species living within a sufficiently restricted geographical area that any member can potentially mate with any other member
Hartl & Clark , 1997
If groups are basic, we should be able to identify them unambiguously
i.e. populations are not “mind independent”
Problem: how do we define the edges?
?
If individuals are basic, then we can define a population by interactions
between individuals
This definition means that populations can still be defined,
even if the edges are vague
This means that we can assign individuals to several populations
... depending on our purpose
We can define populations as groups of interacting individuals, but
how are groups separated?
Groups can become separated by barriers reducing gene flow
The Great Wall of China acts as a barrier to gene flow
Su, H et al. The Great Wall of China: a physical barrier to gene flow?. Heredity, 90, 212 - 219, (2003).
Experiencing different environments can also lead to separation of groups
Populations can become differentiated phenotypically
Bromarv Kerimäki Sotkamo Salla
8090
100
110
120
130
Day
s to
bud
set
Variation Between PopulationsBud set 0.364Allozymes 0.02RFLP 0.02Microsats 0.014
Finland
Using neutral markers to separate populations privileges neutral
genetic data
The methods definewhat a population is
Population 2 Population 3Population 1
What Structure does
Note: number of populations pre-defined
What BAPS does
Population 4Population 3Population 1 Population 2
Population 5Population 6Population 7Population 8
The mathematics behind clustering: The Posterior
Pr(M, P | G) ∝ Pr(M) Pr(P) Pr(G | M, P)
M – Population membershipP - allele frequenciesG - Genotypes( data)
Likelihood(effect of the data)
Prior distributions
The Likelihood for Population Membership
Population 2 Population 3Population 1
Assume Hardy-Weinberg equilibriumIf we move an individual from one population to
another, this changes the likelihood
Allele frequencies
p1,l,a
p2,l,a
p3,l,a
P G∣M ,P =∏Pops∏Loci
∏Alleles
p p , l ,an p , l ,a
The Number of Populations: Goodness of fit is penalised by the dimensionAdd a population:
Pr(P) ↓
“Curse of Dimensionality”
Pr(G | M, P) ↑
Estimating the number of populations depends on the balance between prior and likelihood
Pr(M, P | G) ∝ Pr(M) Pr(P) Pr(G | M, P)
Pr(P) ↓ Pr(G | M, P) ↑
Comment: there is no proof the balance is correct- problem in statistics (AIC, BIC, CIC, DIC, FIC, TIC)
Simulations: Clustering Individuals
5 10 15 20
020
4060
8010
0
Number of Loci
Num
ber o
f Pop
ulat
ions
Defining populations with software
A population is defined using neutral markers
Populations are defined by the amount of data
Prior information very importantarea of developmentallows other factors in
There are also biological reasons to be careful interpreting K.
The population model that we have adopted here is obviously an idealization.
However...clusters may not necessarily correspond to “real” populations.
Pritchard, 2000
Conclusions
“If all you have is a hammer, everything looks like a nail”
Are populations defined by the methods we use?
Human Population Structure
Rosenberg et al. (2002)
93 to 95% of total genetic variation within populations
Identify 6 clusters
Human Population Structure
Manica et al. (2005)
An alternative view: diversity as a cline
The Take Home Message
Populations do not exist
Populations should be defined according to our purpose
Defining populations with dumb software should be done with care
Journal of Negative ResultsEcology & Evolutionary Biology
www.jnr-eeb.org
ReferencesCorander et al. (2003) Bayesian Analysis of Genetic Differentiation
Between Populations. Genetics 163: 367–374.
Gannett (2003) Making Populations: Bounding Genes in Space and in Time. Philosophy of Science, 70: 989–1001.
Hartl & Clark (1997) Principles of Population Genetics
Manica et al. (2005) Geography is a better determinant of human genetic differentiation than ethnicity. Human Genetics in press
Maynard Smith (1998) Evolutionary Genetics
Mayr (1942) Systematics and the Origin of Species
Pritchard et al. (2000) Inference of population structure using multilocus genotype data. Genetics, 155: 945–959.
Rosenberg et al. (2002) Genetic Structure of Human Populations. Science, 298: 2381-2385
Su et al. (2003) The Great Wall of China: a physical barrier to gene flow? Heredity, 90: 212 – 219.
Gannet suggests that populations are “created”, depending on
circumstances
If the appropriate ontology ... treats organisms ... as basic and populations as emergent, ... this challenges population-based approaches that assume that populations exist as mind-independent objects ...
Gannett, 2003