Transcript Phylogenetic Analysis Part 2

Phylogenetic Analysis – Part 2
Spring 2014
Outline
 Why do we do phylogenetics
(cladistics)?
 How do we build a tree?
 Do we believe the tree?
 Applications of phylogenetics (cladistics)
Why do we do phylogenetic analyses?
Evolutionary interpretation of relationships between
organisms.
More reliable method of ascertaining the best hypothesis
for how the organisms diverged from common
ancestors.
Provides a better rationale for circumscribing taxonomic
groups.
Enables testing of multiple hypotheses of relationships
and character evolution based on specific character
transformations.
Willi Hennig
(1913-1976)
German Entomologist (Diptera)
Warren Herbert Wagner
(1920-2000)
American Botanist
(Pteridophytes,
Dendrogrammaceae)
How do we build a tree?
 Data: Types of characters
 Look for synapomorphies
 Assemble the hierarchy of
synapomorphies according to the
principle of parsimony
Parsimony
 The idea that the simplest hypothesis is the
best explanation given the assumptions of the
analysis.
 Also known as Occam’s Razor
 In systematics, we look for the shortest trees
in any given phylogenetic analysis (maximum
parsimony)…
 …even though we know that evolution doesn’t
necessarily proceed in a parsimonious
manner.
Data: type of characters
Character Selection:
Must study each individual character to assure comparisons of
homologous characters, and interpretation of the direction of
character state transformations.
Some sources of characters:
1) morphology (external structure) or anatomy (internal structure)
2) biochemical (photosynthetic pathway, pigmentation pathways,
etc.)
3) chromosome numbers
4) nucleotide sequence data ( a t c g )
Character states
 Presence vs. absence (0 or 1): always binary
(two states)
 Other binary characters (0 or 1): yellow vs.
white flowers; range of measurements
(quantitative)
 Multi-state characters (3 or more states): e.g.,
yellow, white, or pink flowers or a t g c for
nucleotide sequence data
 For binary or multi-state characters, can
hypothesize the direction of evolutionary
change (transformation series)
Pollen wall thickness µm (x axis)
Fig. 2.2
Fig. 2.4
Examples of character state transformations
But how to hypothesize direction?
 We usually have some working knowledge of
the group under study (ingroup).
 We choose a group thought to be closely
related to the ingroup to serve as the outgroup
as a basis of comparison.
 Character states in the outgroup are assumed
to be ancestral (plesiomorphic).
 This allows us to establish the direction of
character state transformation.
Phylogenetics or Cladistics
Dixonia
Dixonia
Cladistics
Attigalea
Dixonia
Ingroup
We need an outgroup, so we
choose the closely related
genus Attigalea (there is
evidence for this from other
studies).
Cladistics: characters
1. Leaf arrangement: alternate (0) or
opposite (1)
2. Leaf midrib: white (0) or green (1)
3. Leaf apex: rounded (0) or spiny (1)
4. Sepals: present (0) or absent (1)
5. Flower color: yellow (0) or blue (1)
6. Subtending floral bracts: absent (0) or
present (1)
Cladistics: characters
1. Leaf arrangement: alternate (0) or
opposite (1)
2. Leaf midrib: white (0) or green (1)
3. Leaf apex: rounded (0) or spiny (1)
4. Sepals: present (0) or absent (1)
5. Flower color: yellow (0) or blue (1)
6. Subtending floral bracts: absent (0) or
present (1)
Cladistics: matrix
Character
1
2
3
4
5
6
Species 1
1
1
0
1
1
0
Species 2
1
0
0
1
0
0
Species 3
1
1
0
1
0
1
Species 4
1
1
0
1
1
0
Species 5
1
1
0
1
0
1
Species 6
1
1
1
1
1
0
Species 7
0
0
0
0
0
0
Species 8
0
0
1
0
0
0
Outgroup (Attigalea)
Ingroup (Dixonia)
Cladistics: matrix
Character
1
2
3
4
5
6
Species 1
1
1
0
1
1
0
Species 2
1
0
0
1
0
0
Species 3
1
1
0
1
0
1
Species 4
1
1
0
1
1
0
Species 5
1
1
0
1
0
1
Species 6
1
1
1
1
1
0
Species 7
0
0
0
0
0
0
Species 8
0
0
1
0
0
0
Cladistics
7
8
2
4
3
5
Sepals absent
Leaves opposite
6
1
Cladistics: matrix
Character
1
2
3
4
5
6
Species 1
1
1
0
1
1
0
Species 2
1
0
0
1
0
0
Species 3
1
1
0
1
0
1
Species 4
1
1
0
1
1
0
Species 5
1
1
0
1
0
1
Species 6
1
1
1
1
1
0
Species 7
0
0
0
0
0
0
Species 8
0
0
1
0
0
0
Cladistics
7
8
2
4
3
5
Midrib green
Sepals absent
Leaves opposite
6
1
Cladistics: matrix
Character
1
2
3
4
5
6
Species 1
1
1
0
1
1
0
Species 2
1
0
0
1
0
0
Species 3
1
1
0
1
0
1
Species 4
1
1
0
1
1
0
Species 5
1
1
0
1
0
1
Species 6
1
1
1
1
1
0
Species 7
0
0
0
0
0
0
Species 8
0
0
1
0
0
0
Cladistics
7
8
2
5
3
4
6
1
Floral bracts
present
Blue flowers
Midrib green
Sepals absent
Leaves opposite
Cladistics: matrix
Character
1
2
3
4
5
6
Species 1
1
1
0
1
1
0
Species 2
1
0
0
1
0
0
Species 3
1
1
0
1
0
1
Species 4
1
1
0
1
1
0
Species 5
1
1
0
1
0
1
Species 6
1
1
1
1
1
0
Species 7
0
0
0
0
0
0
Species 8
0
0
1
0
0
0
Attigalea
Dixonia
purplebracts
floral
fringe
present
midrib green
sepals absent
Length = 5 steps
Cladistics: matrix
Character
1
2
3
4
5
6
Species 1
1
1
0
1
1
0
Species 2
1
0
0
1
0
0
Species 3
1
1
0
1
0
1
Species 4
1
1
0
1
1
0
Species 5
1
1
0
1
0
1
Species 6
1
1
1
1
1
0
Species 7
0
0
0
0
0
0
Species 8
0
0
1
0
0
0
Polymorphism in the outgroup
Homoplasy?
Attigalea
Dixonia
spiny leaf
tip
purplebracts
floral
fringe
present
midrib green
sepals absent
Length = 7 steps
spiny
leaf tip
Do we believe the tree?
 Various programs to generate trees.
 Various measures of statistical support
for the clades and for the characters.
 Can quantify the effects of homoplasy.
 Can test alternate arrangements to
examine the number of steps involved.
Attigalea
Dixonia
spiny leaf
tip
purplebracts
floral
fringe
present
midrib green
sepals absent
Length = 10 steps (7 - 1 + 4)
spiny
leaf tip
Phylogenetic Methodology
1. Selection of taxa to study - Individuals, populations,
species, etc. identified as the units of comparison.
One or more related groups (outgroups) necessary to
“root” the trees.
2. The units under study described for as many
characters as possible for which homology can be
demonstrated or reasonably assumed. Character
states assigned based on variation among the taxa in
the ingroup and outgroup(s) and a priori hypotheses
of the evolutionary direction of changes undergone
by the character (character state polarization) are
generated based on outgroup comparison.
3. A data matrix is assembled by scoring all taxa for all
characters (ideally).
Phylogenetic Methodology
(continued)
4. Using various analytical principles (maximum
parsimony), a cladogram is constructed using
synapomorphic character state changes to determine
the tree topology. Systematists are seeking to define
monophyletic groups (= clades).
5. Evaluation of the statistical confidence in how
robustly the data support the grouping of organisms
into clades is done next.
6. The subsequent grouping and ranking of the
organisms in the resulting clades is then applied to
classification systems or other questions.
Applications of phylogenetics
 Classification
 Biogeography
 Many other possibilities (e.g., disease
tracking, gene annotation)
Cladistics and Classification
- Classifications based on tree topologies…Is it the best
tree? Will new data and new tree topologies
necessitate complete overhaul of classification?
- Remember, systematists would like to define and name
monophyletic groups (clades). Is this always
possible? Must we accept paraphyletic groups?
- Classifications must also be useful for communication.
Do the groups truly represent the best evolutionary
hypothesis, and yet are the classifications useful?
Phylogenetics and Classification
Phylogenetics and Classification
Fig. 2.19A-C
Phylogenetics and Biogeography
Adansonia
(baobab tree)
(Judd et al. 2008)
Fig. 2.22
Phylogenetics
and
Character
Evolution
Phylogenetics
and Conservation
Anomochloa
Grama grass
bamboo
Northern sea oats
bluegrass
Big bluestem
Fig. 19.11