BIO 370 Study Guide - Midterm Guide: Chronogram, Distance Matrix, Transversion
Phylogenetic trees fail to accurately describe: sexual reproduction, hybridization, horizontal
gene transfer.
Phylogenies can indicate evolutionary relationships only, or they can convey info regarding
amount of character change that has occurred along each branch.
-Cladograms: only evolutionary relationships.
-Phylogram: branch lengths indicate amount of sequence change.
-Chronogram: branch lengths rep amount of evolutionary change.
• Oa’s razor: used to hoose the tree ith the least uer of hpothesis of aalog. If
we assume that all life is related through descent, then our simplest interpretation
would be descent due to homology.
• Ingroup – the group of interest, presumably monophyletic
• Outgroup – starting group because we know the natural history of the organism.
• Character mapping: ambiguous
• Parsimony: simplest explanation is the best.
o Problems: have to introduce new term, parsimony informative trait (trait that
unambiguously divides 4 OTUs along a branch)
• Long-branch attraction: due to chance alone, increase in data increases odds of getting
the wrong tree.
o With ore haraters that do’t hage on the internal tree
o To fi this: use odels of eolutio that do’t rel o assigig haraters to
nodes under parsimony, add more taxa to our analysis of ingroup to shorten
branches and increases chances of intersecting long branches, use closest
related outgroup to reduce character change b/w two nodes and a terminal OTU
= improves parsimony and ML. Sample more OTUs which will intersect long
branches and reduce odds of LBA.
o Becomes a problem for parsimony analysis because the character states become
random on certain branches, and for likelihood analysis if we misspecify the
model.
• Distance method: more similar species are more closely related.
o HCA and branches are unrelated to characters and traits, and are unrelated to
character state changes.
o NOT immune to saturation and long-branch attraction.
o Advantage: computational, faster. Can adjust distance matrix by transition-
transversion bias, produce different probabilities of state changes, diff
algorithms may account for some distortions of multi-dimensional space better
than others (more faithfully represent the data), and are highly context-
dependent.
o Disadvantage: lacks underlying evolutionary model, has phonetic approach =
only group species together based on their similarities and not question
underlying historical evolutionary relationships (similarity reflects homology, not
analogy), have to assume that the more DNA sequence differ the more distantly
related species are (might not be the case for species that had rapidly evolved
than others).
find more resources at oneclass.com
find more resources at oneclass.com
