Lecture 3: Overview of phylogenetic comparisons

1
Lecture 3Overview of phylogenetic comparisons
tree-building
2
Phylogenetic trees

• Definition a tree showing evolutionary
  relationships among various biological species or
  other entities that are believed to share a
  common ancestor.
• Features
• Nodes represent most recent common ancestor of
  the descendants
• Branches may correspond to time estimates or
  character substitutions
• Rooted contains a unique node corresponding to a
  shared common ancestor
• Unrooted illustrates relatedness of leaf nodes
  without assumptions of common ancestry

3
An example

• Tree constructed using neighbor joining
• Scale bar corresponds to substitutions per
  nucleotide position
• A variety of Alphaproteobacteria sequences were
  used as outgroups
• Closed circles represent gt 90 and open circles
  gt 70 bootstrap support

Giovannoni et al. 2008
4
Neighbor Joining

• A method used for DNA or protein based trees
• Algorithm requires a distance matrix
• Joins taxa together with similar distance (joins
  the closest neighbors)
• Based on the minimum-evolution criterion, i.e.
  topology providing the least total branch length
  is preferred at each step of the algorithm
• Does not assume all lineages evolve at the same
  rate (no molecular clock assumption)
• Computationally efficient
• Con Trees are often close to optimum, but may
  not be most accurate

5
ARB distance treeing methods

• ARB Neighbor Joining
• Phylip Distance Methods
• Neighbor joining
• UPGMA (average linking clustering)
• Fitch (additive tree model)
• Kitsch (similar to Fitch but assumes molecular
  clock)

Phylip (PHYLogeny Inference Package)
resources http//evolution.genetics.washington.ed
u/phylip.html
6
Maximum Likelihood

• The method searches for the tree with the highest
  probability or likelihood.
• It evaluates a hypothesis about evolutionary
  history in terms of the probability that the
  proposed model and the hypothesised history would
  give rise to the observed data set.
• Statistically well-founded method
• Uses all of the sequence information
• Con requires lots of computational power

7
ARB Maximum Likelihood methods

• FastdnaML computationally slow
• AxML (axelerated maximum likelihood for high
  performance computing) - computationally fast
• RAxML (randomized axelerated maximum likelihood
  for high performance computing) faster yet but
  problematic in ARB
• PHYML DNA
• TREE-PUZZLE (quartet puzzling tree)
• PHYML - Amino acids
• Phylip PROML
• Protein_ML

8
Maximum Parsimony

• Non-parametric statistical method for estimating
  phylogenies
• Preferred phylogenetic tree is the tree requiring
  the least evolutionary change to explain the data
• Con Not statistically consistent

9
ARB Maximum parsimony methods

• Phylip DNAPARS
• Phylip PROTPARS

Phylip (PHYLogeny Inference Package)
resources http//evolution.genetics.washington.ed
u/phylip.html
10
Bootstrapping

• A statistical re-sampling procedure commonly used
  for providing confidence to branches in
  phylogenetic trees
• A measure of repeatability, the probability that
  the branch would be recovered if the taxa were
  sampled again
• Bootstrap percentages are biased, usually towards
  an underestimate of confidence (70 might
  actually represent 95)
• Bootstrapping values are typically presented from
  1000 repeated calculations

11
Tutorial 3 objectives

• Add aligned sequences to the ARB database tree
• Refine sequence alignments
• Create a phylogenetic tree in ARB
• Export trees from ARB
• Create trees from web-based resources (RAxML tree)