Lecture 11 - Phylogenetic Trees Flashcards
what are the four steps of phylogenetic data analysis
- selection of sequences
- MSA
- tree building
- tree evaluation
what two things evolve which complicates molecular evolutionary studies
species and genes
define: additive trees
trees that attempt to preserve the sum of distances between terminal elements within the tree
describe distance-based tree methods
convert alignments into distances then use this to infer a tree
describe character-based tree methods
analyze individual character states and specific positions in the alignment and infer a tree based on observed changes
what is the main idea of UPGMA (unweighted pair group method with arithmetic mean)
at each iteration, the two closest groups are combined to form one new group; continue until there is only one group
what is the result of UPGMA
a rooted, non-additive and ultrametric tree
ultrametric - all leaves are the same distance from the root
what is the idea of neighbor-joining (NJ)
minimize the total length of branches on the tree
what do you start with for NJ
a star-shaped tree + pairwise distance matrix
what are the steps of NJ
- join pair with smallest sum of branch lengths with a new node
- regard the pair as a single OTU and do pairwise comparisons again
- repeat until the entire tree is generated
what is the result of NJ
a tree that accounts for variation in branch lengths and fits well with statistical models of evolution
define: bootstrapping
computational technique for estimating the confidence level of a phylogenetic hypothesis
what are bootstrap values calculated for
individual nodes (and/or branches)
what value is often used as a bootstrap value threshold
> 70%
how is bootstrapping done for trees
- divide the MSA into a set of N ordered sites and randomly choose N sites with replacement
- recalculate trees (100 or 1000 replicates is common) and determine the frequency of each node within the replicated
what types of tree methods are bootstrapping usually used for
rapid phylogenetic methods such as NJ
what do parsimony methods look for
the fewest number of changes in an evolutionary tree
what is a tree with the fewest number of changes equivalent to
a tree of minimum length
what is the starting point of parsimony
MSA
what are the two types of sites within an MSA defined for parsimony
- informative
- uninformative
what is an uninformative site
invariant sites
what defines an informative site
a site with has at least two different nucleotides that are present two or more times
what is the general approach to building trees for comparison
- build a starting tree using a fast method
- make changes to that tree by moving branches or nodes
- over time, keep trees that have an improved score
how does NNI work
- each internal branch is connected to four groups within a tree
- change how an internal branch is connected
- if the score is better, keep the new tree
