Tree reliability

Question 1

With tree length frequency distributions what do skewed distributions indicate

Answer 1

stronger phylogenetic signal

Question 2

In tree length frequency distributions what tree is considered more reliable for the lower of these two data sets?

Answer 2

the shortest tree?

Question 3

What is the formula for consistency index

Answer 3

CI = (m/s)

–> m = minimum number of steps (number of states - 1)
–> s = number of steps in the tree (for that character)

–> for the whole tree you do sum of m/sum of s for each character

Question 4

What does the maximum value of one mean for a Consistency index

Answer 4

this means no homoplasy is observed

Question 5

what is the problem with the consistency index

Answer 5

• autapomorphies inflate the CI but are not informative

Question 6

What is the definition of the retention index

Answer 6

fraction of potential synapomorphies retained as synapomorphies on the tree

Question 7

what is the formula for retention index

Answer 7

RI = (g-s)/(g-m)

–> m = minimum number of steps (number of states - 1)
–> s = number of steps in the tree (for that character)
–> g is the maximum number of steps = number of steps on a polytomy

Question 8

What is the better estimate of the efficiency with which the tree explains the data: RI (retention index) or CI (consistency index)

Answer 8

RI because autapomorphies do not inflate it unlike for CI

Question 9

Describe consensus trees

Answer 9

• the goal is to summarize information from rival trees
• can have a strict consensus tree where the tree contains the clades that occur in ALL the rival trees
• can have majority rule consensus trees where clades included are those that are found in a majority of the trees

Question 10

Describe Bremer support

Answer 10

the difference in number of steps between the length of the most
parsimonious tree(s), and the length of the most parsimonious tree
that does not contain a particular clade (node, branch)

Question 11

Describe Bootstrap support

Answer 11

this method involves re sampling the data with replacement. You draw one character at random and then put it back and draw again until you have drawn the same amount of characters you originally had in your matrix (so characters can be repeated). You save the best tree for each replicate and then find the majority tree of all replicates. The proportion of trees supporting a clade is the bootstrap proportion

Question 12

How does Jackknife support work

Answer 12

• this is also a resampling procedure like bootstrapping but here you subsample within a cycle?
• it resamples a proportion of the characters WITHOUT replacement (unlike bootstrapping)
• here the best tree from each replicate is saved, and a majority rule consensus is obtained from all the trees, the proportion of trees supporting a clade is the jackknife proportion

Question 13

Describe posterior probability

Answer 13

• used with Bayesian analysis
• the posterior probability of a tree is the conditional probability that the tree is correct, taking into account new evidence an the prior probability
• generates millions of trees using a kind of random walk simulation technique (MCMC)
• compiles the trees into a single consensus tree

Question 14

Compare Bayesian values with bootstrap values

Answer 14

Bayesian values are higher

Question 15

Describe multispecies coalescent probability

Answer 15

• this estimates gene tree-species tree congruence
• coalescent theory is a model of how gene variants sampled from a population may have originated from a common ancestor