Phylogenetic Tree Building

One-Sentence Definition

Phylogenetic tree building infers evolutionary relationships among sequences by fitting a branching model to aligned characters (nucleotides, amino acids, or alleles).

Simple Explanation

Turn sequence differences into a family tree, then judge how much to trust each branch.

Detailed Scientific Explanation

MethodIdeaNotes
Distance (NJ)Cluster by pairwise distancesFast, less accurate
Maximum parsimonyFewest changesSensitive to long-branch attraction
Maximum likelihood (IQ-TREE, RAxML)Best fit under substitution modelStandard for genomics
Bayesian (MrBayes, BEAST)Posterior distributionsAdds time/priors — see Phylodynamics

Practical points:

  • Substitution model selection (e.g., GTR+G) matters
  • Support values: bootstrap / ultrafast bootstrap / posterior probabilities
  • Rooting: outgroup, midpoint, or molecular clock
  • Recombination distorts trees — mask with Gubbins/ClonalFrameML in recombinogenic species

Mechanism

Alignment → model → search tree space to maximize likelihood → assess support by resampling.

Clinical Importance

  • Trees underlie outbreak claims; poor support ⇒ weak transmission conclusions

Research Importance

  • Taxonomy (Carl Woese lineage of thought), evolution of resistance lineages

Diagnostic Relevance

  • Public-health genomics reporting

AMR Relevance

  • Distinguishes vertical spread of a resistant clone from repeated horizontal acquisition

Active Recall Questions

  1. Why is maximum likelihood preferred over neighbor-joining for genomics?
  2. What do bootstrap values express?
  3. Why mask recombination before tree building?

Connections