Flashcards in Lecture 19 Deck (21):
1
Outgroup
-lacks all characters that are seen in other branches
-character state is 0
2
Parsimony (2 concepts)
-methodological
-ontological
3
Methodological Parsimony
-Ockham's Razor
-best hypothesis is simplest i.e. makes fewest assumptions
4
Ontological Parsimony
-nature prefers the simplest course of evolution (not necessarily true; what if nature doesn't always prefer the fewest number of changes? homoplasies do exist-computers let you know how many exist in each tree)
-thus the best tree requires the fewest changes i.e. the fewest homoplasies
5
Determining proper tree
-add more characters
-the longer the think is the less likely it is to be right
-want fewest number of informative steps
-want to find tree with shortest length, and reduce number of homoplasies
6
Automorphy
-only present in one clade
-not informative
-not used in analysis at end
-homoplasy on these trees is a characteristic that arises twice
7
Molecular Sequences
-loses relationship with extinct things but gains ability to see relationship between very different organisms like ears of corn or mushrooms as opposed to doing it morphologically
-genes don't care about morphology
-don't have to have any morphological characters in common to get relatedness
8
Parsimony in Cladistics
-use only synaptomorphies
-character present in only one lineage not used in cladistics
-lose some data when you do it cladistically
-will use different data set depending on what method of analysis you use
9
Why is there so much homoplasy in molecular data than in morphological data?
-mutation doesn't always cause a morphological change
-more fundamental reason: only four possible bases for each position so you need a lot of sequence when you're using this method because the homoplasy rate is so high
10
How many bp in human genome?
-3 billion
11
How do you estimate divergence times?
-fossil record
12
Unrooted Tree
-don't know topology of relationships seen in the data
-used to give you a rather large scale map
13
Ordering States of Features for Analysis
1. select features
2. decide what are primitive and derived states for each feature
3. generate data set
4. application to cladogram inference
14
Polarity Assignment
-Has to be determined empirically, as in examples; fossil data can help
-states can be assigned numbers for easy analysis
-e.g. 0 for primitive state, 1 for derived. If a second derived state exists can be called 2.
-with that, a table of taxa and character states can be prepared
15
When Building a Simple Cladogram
-0 means trait is absent 1 means trait is present
-computer doesn't build a tree from the data it matches the data to all possible trees
16
Inferring Trees
-Number of taxa and characters increases the complexity
-with number of taxa, number of possible trees increases fast
17
Phylogenetic Methods Used on Gene Sequence Data
-lots of data
-need homologous genes from different taxa e.g. ribosomal genes-universal
-evolves by substitution of bases
-first step sequence genes
-second step align gene sequence from taxa
-third step apply tree algorithms can use a few distinct methods
18
Methods of Alignment and Inference
-Distance: use all differences
-Parsimony: use only synapomorphy (cladistic)
-look at picture in notes
19
Phylogenetic Trees
-not all are cladograms
-in molecular phylogeny what are called "distance trees" also are used
-branch length scan differ (a reflection of evolutionary rate along the branch)
-evolution of genes does not necessarily have a uniform rate
20
Value of DNA Sequence Phylogeny
-huge amount of data
-can bridge taxa that share no morphology
-can estimate divergence times
21