Methods of Molecular Phylogeny Reconstruction & Evaluation

(A) Phenetics:   given a metric for a matrix of pairwise differences (e.g., DNA sequence), which clustering minimizes the overall length of the tree? [e.g, NJ: Neighbour-Joining]

(B) Cladistics:   given the distribution of various character states (A,C,G,T), which branching order requires the smallest number of changes? [e.g., MP: Maximum Parsimony]

(C) Maximum-Likelihood:    Given a particular set of parameters for a model of (molecular) evolution,  which of the many unlikely outcomes is least unlikely (maximally likely)?

(D) Cladograms vs Phylograms: diagrams of evolutionary relationships may be cladograms that simply depict branching order (A, B, & C), or phylograms that show the amount of change occuring along each branch (D)

Bootstrap Analysis: Given a finite data set, re-sample it as though it were a random sample of an infinite data set. In what % of cases are the same groups observed? [(A-C) numbers above branches, applicable to groups to the right]

Phylogeny Reconstruction methods

Neighbor-Joining (A), Maximum Parsimony (B), and Maximum Likelihood (C) cladograms, with bootstrap support; parsimony phylogram, with branch lengths (D)


All text material © 2007 by Steven M. Carr