Learning Objectives
360 words
2 min read
Visual companion
Python
Type and operator map
Python Week 1: the first filter for runtime behavior
View
Revision summary
What this note is really saying
Short form
# Learning Objectives - Understand phylogenetic tree concepts - Compare distance-based and character-based methods - Interpret bootstrap support values ## 1. Tree Concepts Leaves = existing species/sequences, internal nodes = common ancestors, branches = evolutionary time/length, topology = branching pattern.
Learning Objectives
- Understand phylogenetic tree concepts
- Compare distance-based and character-based methods
- Interpret bootstrap support values
1. Tree Concepts
Leaves = existing species/sequences, internal nodes = common ancestors, branches = evolutionary time/length, topology = branching pattern.
2. Tree Building Methods
Distance-Based (UPGMA, Neighbor-Joining):
- Calculate pairwise distances between sequences
- Group closest pairs, recalculate distances to new groups
- Repeat until complete Fast, but can lose information. UPGMA assumes constant evolutionary rate (molecular clock). Maximum Parsimony: Find tree requiring fewest evolutionary changes (mutations). Minimize total number of changes. NP-hard. Prone to long-branch attraction. Maximum Likelihood: Find tree with highest probability of producing observed sequences given an evolutionary model. Statistically rigorous, computationally intensive. Preferred method. Bayesian (MrBayes): Sample trees from posterior distribution using MCMC. Provides branch support probabilities directly.
3. Tree Evaluation
Bootstrapping: Randomly resample alignment columns (with replacement), rebuild tree, see if same branches appear. >70% bootstrap support = strong evidence. >95% = very strong.
Q1: What is the difference between distance and character-based methods?Distance: convert sequences to pairwise distances, then cluster. Fast but loses character-level information. Character-based: use actual sequence characters (parsimony, likelihood). More accurate but slower. Q2: What does bootstrapping measure?Confidence in tree branches. Resample alignment columns with replacement 100-1000x. Branch support = % of bootstrap trees showing that branch. >70% = strong support. Q3: What is maximum parsimony?Tree requiring fewest evolutionary changes. Simple assumption: prefer simpler explanation. Computationally NP-hard. Can suffer from long-branch attraction. Q4: What is the molecular clock hypothesis?Sequences evolve at constant rate over time. If true, branch lengths proportional to time. UPGMA assumes this. Many datasets violate this (rate variation among lineages). Q5: What is long-branch attraction?Rapidly evolving lineages (long branches) incorrectly grouped together by parsimony due to convergent changes. Maximum likelihood less susceptible. Q6: What is the difference between rooted and unrooted trees?Rooted: identifies common ancestor, shows direction of evolution. Unrooted: only shows relationships, not direction. Outgroup species can be used to root trees. Join Discord PreviousHidden Markov ModelsNextMotif Finding & Regulatory Genomics