Skip to content

Chapter 9: Phylogenetics: Understanding Evolutionary Relationships

📥 Download Lecture Slides (PPTX)

9.1 The Tree of Life

"Nothing in biology makes sense except in the light of evolution." - Theodosius Dobzhansky.

Illustration of a Phylogenetic Tree

For an interactive view of tree manipulation and collapsing, open the interactive demo: Interactive Phylogenetic Tree

Phylogenetics is the study of evolutionary relationships among groups of organisms. We represent these relationships using a Phylogenetic Tree.

  • Leaves (Tips): The current species or sequences we are analyzing.
  • Nodes: The hypothetical common ancestors.
  • Branches: The lines connecting nodes. The length often represents time or amount of genetic change.
  • Root: The oldest point in the tree, representing the common ancestor of all species in the tree.

9.2 Multiple Sequence Alignment (MSA)

Before you can build a tree, you must align your sequences. Since we are comparing more than two, we use Multiple Sequence Alignment.

Imagine stacking 10 DNA sequences on top of each other. You need to insert gaps so that the "columns" of the stack align (e.g., all the ancestors' Adenines line up).

  • Tools: ClustalW, MUSCLE, MAFFT.

9.3 Building the Tree

There are two main approaches to building trees:

  1. Distance-Based (e.g., Neighbor-Joining):

    • Calculate the percent difference between every pair of sequences.
    • Group the two most similar ones, then the next, until the tree is finished.
    • Pros: Very fast. Cons: Less accurate for distant relationships.

  2. Character-Based (e.g., Maximum Likelihood, Bayesian):

    • Uses statistical models of how DNA mutates (e.g., transitions are more common than transversions).
    • Calculates the probability of the tree given the data.
    • Pros: Highly accurate. Cons: Computationally expensive (slow).

9.4 Bioinformatics in Action: Drawing a Tree

Biopython can parse tree files (often in "Newick" format) and visualize them.

from Bio import Phylo
from io import StringIO

# A sample tree in Newick format (usually this comes from a file)
# ((Raccoon, Bear), ((Sea_Lion, Seal), ((Monkey, Cat), Weasel)), Dog);
tree_data = "((Raccoon:10, Bear:10):5, ((Sea_Lion:8, Seal:8):4, ((Monkey:15, Cat:15):3, Weasel:12):2):3, Dog:20);"

# Read the tree
handle = StringIO(tree_data)
tree = Phylo.read(handle, "newick")

# Draw it in ASCII art (perfect for the terminal!)
Phylo.draw_ascii(tree)

Output: 

                                            __________________ Raccoon
  _________________________|
 |                         |__________________ Bear
 |
 |                                            ________ Sea_Lion
 |                       ____________________|
 |                      |                    |________ Seal
 |______________________|
 |                      |                     ________________ Monkey
 |                      |____________________|
 |                                           |________________ Cat
 |                                           |
 |                                           ____________ Weasel
 |
 |____________________ Dog

Summary

Phylogenetics allows us to reconstruct the history of life. We start with an MSA, choose a method (like Maximum Likelihood), and produce a tree that visualizes the evolutionary distance between species.

9.5 Model Selection, Tools, and Robustness

Modern phylogenetic practice emphasizes model selection, statistical support, and reproducibility.

  • Model selection: Use tools like ModelFinder (built into IQ-TREE) to select substitution models that best fit your MSA.
  • Tree inference: Popular, fast, and reliable tools include IQ-TREE and RAxML-NG for Maximum Likelihood trees; MrBayes for Bayesian inference.
  • Support values: Perform bootstrapping (standard or ultrafast bootstrap in IQ-TREE) to quantify node support.
  • Visualization: Use ETE3, FigTree, or interactive services like iTOL to visualize annotated trees.

Example: running IQ-TREE with model selection and ultrafast bootstraps:

# Model selection + ML tree + ultrafast bootstrap (1000 replicates)
iqtree2 -s alignment.fasta -m MFP -B 1000 -T AUTO

Practical tips:

  • Inspect the MSA for poorly aligned regions and trim them (e.g., trimAl) before building trees.
  • For large datasets, consider partitioning or using gene/species tree reconciliation methods.
  • Share alignment, trees, and commands (e.g., README or workflow file) so results are reproducible.