Lecture 11

Question 1

Molecular clock hypothesis

Answer 1

DNA and protein seq evolve at rate that is relatively constant over time and among diff organisms-> genetic differences btwn 2 spp are proportional to time

Question 2

Zuckerkandl and Pauling

Answer 2

1st to suggest molecular clock; # of amino acid differences in hemoglobin; 1962

Question 3

Motoo Kimura’s Neutral Theory of Evolution

Answer 3

1968; neutral mutations would either become fixed in a pop or lost through stochastic genetic drift
-rate at which neutral mutations become fixed (substitution rate) is equivalent to rate of appearance of new mutations in each member of pop (mutation rate)

Question 4

Index of dispersion (general)

Answer 4

Kimura (early 1970s)

• if all spp diverged at same time and had same substitution rate-> variance-to-mean ratio of # of subs among lineages would not be sig different than 1
• is there more rate variation between lineages than expected under poisson

Question 5

Index of dispersion (understanding)

Answer 5

I = variance/mean
I<1: clock
I = 1: null
I often called R(t)

Question 6

Relative rates test

Answer 6

• used to estimate the difference in # of subs btwn 2 closely related taxa in comparison to 3rd outgroup
• does not require knowledge of divergence times of taxa

Question 7

Why can’t outgroups be too distantly related in relative rates test?

Answer 7

Greater problem of multiple hits and smaller the impact any difference in rate will have on distance measurements

Question 8

strict molecular clocks requirements

Answer 8

• all internodes have equal duration
• all branches have equal sub rate
• all tips same # of time units away from root
• the observed # of subs is the same for all descendants of a given node
• expected # of subs per site is same for all branches

Question 9

how to test for strict molecular clock

Answer 9

likelihood ratio test:
˚null hypothesis (H0):constrained branch lengths
˚H1: each branch allowed to vary independently; (n-2) additional parameters
˚-2logL = 2(logL0 - logL1)
˚significance approximated with X^2 distribution (with n-2 deg of freedom)

Question 10

relaxed clocks

Answer 10

-allows the evolutionary rate to “evolve” over time, assuming it is linked to other evolving biological traits

Question 11

Rate-smoothing

Answer 11

-relaxed molecular clock method; rates of neighboring branches of tree are constrained to be similar (r8s)

Question 12

Thorne’s Bayesian Relaxed Clock

Answer 12

• molecular rates among lineages allowed to vary in autocorrelated manner(rate of descendants depends on common ancestor)
• rate in each branch drawn a priori from parametric dist whose mean is a function of rate on parent branch
• integrates over range of possible rates instead of fixed value for each branch

Question 13

Bayesian MCMC

Answer 13

co-estimate of phylogeny and divergence times using probalistic calibration priors (instead of point calibrations); rates allowed to vary in uncorrelated manner (lognormal or exponential)

Question 14

r8s

Answer 14

• ML methods to experimental semiparametric and nonparametric methods to relax stringency
• input is phylogenetic tree with estimated branch lengths and a calibration point

Question 15

shortcomings of r8s

Answer 15

requires at least 1 node to be fixed in time; uncertainty in phylogeny not accounted for

Question 16

What types of calibrations provide time?

Answer 16

• constant rates of molecular change (fixed sub rate)
• fossils
• geologic info (max or min)