Jacques'van'Heldenulb'ac'be

1
Phylogeny

• Bioinformatics

2
Species trees versus molecule tree

• A species tree aims at representing the
  evolutionary relationships between species.
• A molecule tree represents the evolutionary
  history of a family of related molecules (genes,
  proteins).
• Species trees and gene trees are generally
  related ...
• Species tree can be inferred from various
  criteria, including the history of carefully
  chosen molecules.
• ... but not identical.
• A molecular family can contain several copies in
  the same species (in-paralogs), due to gene
  duplications.
• Some molecules can be transferred horizontally
  between species.
• Due to combinations of duplications-divergences,
  the tree of a given gene may be inconsistent with
  the species tree.
• Illustration Figure 7.3 from Zvelebil and Baum.

Source Zvelebil, M.J. and Baum, J.O. (2008)
Understanding Bioinformatics. Garland Science,
New York and London.
3
Tree reconciliation
Source Zvelebil, M.J. and Baum, J.O. (2008)
Understanding Bioinformatics. Garland Science,
New York and London.
4
Concept definitions from Fitch (2000)

• Discussion about definitions of the paper
• Fitch, W. M. (2000). Homology a personal view on
  some of the problems. Trends Genet 16, 227-31.
• Homology
• Owen (1843).  the same organ under every variety
  of form and function .
• Fitch (2000). Homology is the relationship of any
  two characters that have descendent, usually with
  divergence, from a common ancestral character.
• Note character can be a phenotypic trait, or a
  site at a given position of a protein, or a whole
  gene, ...
• Molecular application two genes are homologous
  if diverge from a common ancestral gene.
• Analogy relationship of two characters that have
  developed convergently from unrelated ancestor.
• Cenancestor the most recent common ancestor of
  the taxa under consideration
• Orthology relationship of any two homologous
  characters whose common ancestor lies in the
  cenancestor of the taxa from which the two
  sequences were obtained.
• Paralogy Relationship of two characters arising
  from a duplication of the gene for that
  character.
• Xenology relationship of any two characters
  whose history, since their common ancestor,
  involves interspecies (horizontal) transfer of
  the genetic material for at least one of those
  characters.

Analogy Homology Paralogy Xenology or not
(xeonologs from paralogs) Orthology Xenology or
not
5
Exercise

• On the basis of Fitchs definitions (previous
  slide), qualify the relationships between each
  pair of genes in the illustrative schema.
• P paralog
• O ortholog
• X xenolog
• A analog

• Orthologs can fomally be defined as a pair of
  genes whose last common ancestor occurred
  immediately before a speciation event (ex a1 and
  a2).
• Paralogs can fomally be defined as a pair of
  genes whose last common ancestor occurred
  immediately before a gene duplication event (ex
  b2 and b2'). Source Zvelebil Baum, 2000

6
Exercise

• Example B1 versus C1
• The two sequences (B1 and C1) were obtained from
  taxa B and C, respectively.
• The cenancestor (blue arrow) is the taxon that
  preceded the second speciation event (Sp2).
• The common ancestor gene (green dot) coincides
  with the cenancestor
• -gt B1 and C1 are orthologs

• Orthologs can fomally be defined as a pair of
  genes whose last common ancestor occurred
  immediately before a speciation event.
• Paralogs can fomally be defined as a pair of
  genes whose last common ancestor occurred
  immediately before a gene duplication event.
• Source Zvelebil Baum, 2000

7
Exercise

• Example B1 versus C2
• The two sequences (B1 and C2) were obtained from
  taxa B and C, respectively.
• The common ancestor gene (green dot) is the gene
  that just preceded the duplication Dp1.
• This common ancestor is much anterior to the
  cenancestor (blue arrow).
• -gt B1 and C2 are paralogs

• Orthologs can fomally be defined as a pair of
  genes whose last common ancestor occurred
  immediately before a speciation event.
• Paralogs can fomally be defined as a pair of
  genes whose last common ancestor occurred
  immediately before a gene duplication event.
• Source Zvelebil Baum, 2000

8
Solution to the exercise

• On the basis of Fitchs definitions (previous
  slide), qualify the relationships between each
  pair of genes in the illustrative schema.
• P paralog
• O ortholog
• X xenolog
• A analog

9
Cladistics, cladograms and clades

• Cladistics
• (Greek klados branch) is a branch of biology
  that determines the evolutionary relationships
  between organisms based on derived similarities
  (source Wilkipaedia).
• Cladogram
• tree-like drawing, usually with binary
  bifurcations, representing one evolutionary
  scenario about divergences between species or
  sequences.
• Clade
• Any sub-tree of a cladogram.
• Note branch lengths to not reflect evolutionary
  time.

10
Phylogram

• Phylogram tree-like structure representing an
  evolutionary scenario, and including
• the events of divergence between species or
  sequences
• the evolutionary time between each species and
  the divergence events.

11
Molecular clock

• The "molecular clock" hypothesis (left tree)
  assumes that rates of evolution do not vary
  between branches. All leaf nodes are thus aligned
  vertically.
• This hypothesis is not always valid
• in some cases, two genes can diverge from a
  common ancestor, but one of them may have
  diverged faster than the other one. This is a
  rather classical mechanism of evolution a
  duplication creates some redundancy, and one copy
  of the gene will evolve whereas the other one
  retains the initial function.

Ultrametric tree (with clock) (e.g. UPGMA)
Without clock (e.g. neighbour-joining)
12
Phylogenetic inference from sequence comparison
Unaligned sequences

• Alternative approaches
• Maximum parsimony
• Distance
• Maximum likelihood

Sequence alignment
Aligned sequences
strong similarity ?
many (gt 20) sequences ?
Maximum parsinomy
no
yes
Source Mount (2000)
13
Maximum parsimony

• For each column of the alignment, all possible
  trees are evaluated and the tree with the
  smallest number of mutations is retained
• The trees which fit with the highest number of
  columns are retained
• The program can return several trees

Adapted from Mount (2000)
14
Maximum parsimony example

• Parsimony tree calculated from a multiple
  alignment of the E.coli proteins containing a
  lacI-type HTH domain
• Left text representation (protpars output)
• Bottom right visualized with njplot (in the
  ClustalX distribution)

-----------CYTR_ECOLI ------------------
--------6 ! !
--------EBGR_ECOLI !
-13 !
! -----CSCR_ECOLI !
-12 !
! --IDNR_ECOLI !
--5 !
--GNTR_ECOLI --4 !
!
-----MALI_ECOLI ! !
-10 ! !
! ! --TRER_ECOLI ! !
--------------9 -14 ! ! !
! --YCJW_ECOLI ! !
! ! ! ! !
--------LACI_ECOLI !
--------------8 --2 !
--FRUR_ECOLI ! !
! -------15 ! ! !
! --RAFR_ECOLI ! !
----------11 ! !
! -----ASCG_ECOLI ! !
-----7 --1 !
! --GALS_ECOLI ! !
--3 ! !
--GALR_ECOLI ! !
! -----------------------------------------RBSR_
ECOLI ! -----------------------------------
---------PURR_ECOLI remember this is an
unrooted tree! requires a total of 4095.000
15
Maximum parsimony - drawbacks

• Number of trees to evaluate increases
  exponentially with the number of sequences.
• Assumes that all sequences evolved at the same
  rate (molecular clock hypothesis).
• Only works for well conserved sequence families.

16
Phylogenetic inference from sequence comparison
Unaligned sequences

• Alternative approaches
• Maximum parsimony
• Distance
• Maximum likelihood

Sequence alignment
Aligned sequences
strong similarity ?
many (gt 20) sequences ?
Maximum parsinomy
no
yes
no
no
clear similarity ?
Distance
yes
Source Mount (2000)
17
Distance method

• Starting from a multiple alignment, calculate the
  distance between each pair of sequences
• Calculate a tree which fits as well as possible
  with the distance matrix
• branch lengths should correspond to distances
• rooted or unrooted
• Several methods can be used for calculating a
  tree from the distance matrix.
• Fitch-Margoliah
• Neighbour-Joining
• UPGMA

Aligned sequences
Distance calculation
Distance matrix
Tree calculation
Tree
18
Distance matrix

• The distance matrix indicates the distance
  between each pair of sequence.
• The matrix is symmetrical, and the diagonal only
  contains 0s.

19
Trees
Rooted tree
Unrooted tree
Unrooted tree

• The distance between two nodes is the sum of
  lengths of the branches between them

20
Methods for calculating trees from a distance
matrix

• It is usually not possible to find a tree whose
  branch length fit with all the values of the
  distance matrix.
• Several approaches exist to calculate a tree
  which approximates the distances.
• The Fitch-Margoliah method minimizes the sum of
  squares between distances in the matrix and
  distances in the tree.
• The Neighbour-Joining (NJ) method minimizes the
  sum of branch lengths for the resulting tree.
  This methods does not assume a molecular clock
  it is thus appropriate when some proteins
  sequences have evolved faster than some other
  ones. It returns an unrooted tree.
• The Unweighted Pair-Group Method by arithmetic
  Averaging (UPGMA) clusters the sequences by order
  of distance in the distance matrix. This method
  relies on the assumption of evolutionary clock,
  and it produces a rooted tree.

21
Example of phylogenetic tree

• This tree was obtained with the Neighbour-Joining
  method (implemented in ClustalX).
• The drawing was obtained with njplot (part of the
  ClustalX package)
• Each branch of the tree is labelled with the
  distance.

22
Distance-based methods for calculating trees in
the package PHYLIP

• Summary of the methods for calculating a tree
  from a distance matrix.

23
Bootstrapping

• In some cases, the data does not allow to infer
  phylogeny
• To assess the reliability of the inference, one
  can apply the bootstrap method
• Given an alignment of n sequences and p columns,
  one performs a random selection of p columns,
  with replacement. Some columns can thus be
  selected multiple times, whilst some others are
  not selected at all.
• Calculate a tree with the sampled columns.
• Repeat many (e.g. 1000) times, and check whether
  the same branches occur frequently (e.g. gt 70).

24
Phylogenetic inference from sequence comparison

• Alternative approaches
• Maximum parsimony
• Distance
• Maximum likelihood

Unaligned sequences
Sequence alignment
Aligned sequences
strong similarity ?
many (gt 20) sequences ?
Maximum parsinomy
no
yes
no
no
clear similarity ?
Distance
yes
no
Maximum likelihood
Source Mount (2000)
25
Practicals with phylogeny.fr
26
Phylogeny.fr

• http//www.phylogeny.fr
• Offers a user-friendly interface to run all the
  steps for inferring phylogeny from a set of
  unaligned sequences.
• Completely automated workflow or user-specified
  parameters.
• Alternative methods for each step of the
  workflow.
• Results are exported in multiple formats
  (convenient for using them with other programs).
• Results can be displayed immediately (for fast
  programs) or sent by email (slow programs).

27
Phylogeny.fr sequence input

• The one click option only requires for you to
  enter a set of sequences, and click on the
  submit button.

28
Phylogeny.fr work flow

• At each step of the workflow, you can
• Check the parameters used for the analysis
• Choose alternative parameters (advanced use)
• Export the intermediate and final results in a
  variety of formats, which can then be opened in
  other programs.

29
Phylogeny.fr - alignment result
30
Phylogeny.fr - phylogenic tree in text format
31
Phylogeny.fr - Phylogram (various output formats
are supported)
32
Phylogeny.fr - display options
33
Phylogram with an outgroup added (Bacillus) but
not correctly rooted (midpoint grouping)
34
Cladogram incorrectly rooted (midpoint)
35
Phylogram rooted with an outgroup
36
Further reading
37
Further reading

• Textbooks
• Zvelebil, M.J. and Baum, J.O. (2008)
  Understanding Bioinformatics. Garland Science,
  New York and London.
• Mount, M. (2001) Bioinformatics Sequence and
  Genome Analysis. Cold Spring Harbor Laboratory
  Press, New York.
• Pevzner, J. (2003) Bioinformatics and Functional
  Genomics. Wiley.
• all his teaching material on http//pevsnerlab.k
  ennedykrieger.org/bioinfo_course.htm

38
Supplementary material
39
PHYLIP flowchart
Distance calculation protdist dnadist
Bootstrapping seqboot
aligned sequences
distance matrix
Parsimony protpars dnapars
Branch-and-bound dnapenny
Maximum likelihood dnaml protml
UPGMA neighbor (rooted)
Fitch-Margoliash fitch (unrooted) kitsch (rooted)
Neighbor -joining neighbor
tree
consense
retree
Tree drawing drawgram
Tree drawing drawtree
drawing of rooted tree
drawing of unrooted tree
40
Taxonomy of bacteria having a gene metA (August
2004)
Bacteria
Bacillales
Bacillaceae
Bacillus
Firmicutes
Clostridia
Clostridiales
Clostridium
Lactococcus
Streptococcus
Brucella
Alpha subdivision
Rhizobiaceae group
Rhizobium
Sinorhizobium
Proteobacteria
Epsilon subdivision
Campylobacter group
Campylobacter
Escherichia
Enterobacteriaceae
Salmonella
Gamma subdivision
Yersinia
Vibrionaceae
Vibrio
Thermotogae
Thermotogae (class)
Thermotogales
Thermogata
41
Tree nomenclature

• Node
• Leave
• Internal branch
• External branch

42
Alignment methods
Source Zvelebil, M.J. and Baum, J.O. (2008)
Understanding Bioinformatics. Garland Science,
New York and London.
43
Evolutionary model
Source Zvelebil, M.J. and Baum, J.O. (2008)
Understanding Bioinformatics. Garland Science,
New York and London.