Snap shot of different approaches to Tree Building
Similarity/Difference Methods that use clustering algorithms
Repeatable – always results in the same “best” tree
Less computationally intensive – suitable for large # of taxa
UPGMA (EvoBeaker Dogs)
Assumes equal rate of evolution – simple clustering algorithm (average
distances)
Neighbor-joining method
Doesn’t assume equal rate of evolution – slightly more complex clustering
algorithm
Maximum parsimony
Saves the shortest trees (there may be more than one)
Maximum Likelihood
Given a model of how evolution is thought to proceed and the data on hand, what is the
most likely tree?
Baysesian methods
What is the probability of observing a hypothesize tree (or trees) given a model of
evolution and the data on hand