Lecture 11: Phylogenetics II

Clade

• group of organisms that share a common ancestor and includes all descendants

Sister Taxa

• taxa derived from the same common ancestor (node)

Polytomy

• unresolved relationship between 3+ taxa

Monophyletic

• includes the most recent common ancestor and all descendants

  • ideally, organisms are grouped according to pattern of descent, and only monophyletic groups should be named

  • BUT not all groups in practice are monophyletic

Paraphyletic

• includes the most recent common ancestor but not all descendants

• most traditional groups are paraphyletic

  • ex. dinosaurs; Dinosauria should include birds

  • ex. reptiles: Reptilia should also include birds

Polyphyletic

• does not include the most recent common ancestor of all members of a group

• group has at least 2 separate evolutionary origins

Rooted Tree

• common lineage indicated from base of tree

  • direction indicates the passage of time

• in theory, it is possible to root a tree along any branch

  • different roots provide different sequences of branching events

  • one tree will therefore be more correct than others

Unrooted Tree

• does not indicate (fully) the direction of the passage of time

• unclear which internal node is the most ancestral

Root

• point on a tree representing the earliest time in the evolutionary history of the taxa included

  • hypothetical common ancestor to all taxa in the tree

• root trees via outgroups

Model Based Phylogeny

• parsimony analysis does not incorporate statistical model or evolutionary change

  • best tree = one with the fewest changes

• model based methods incorporate probability models of how characters change over time

  • calculates probability of a change occuring in a given branch

• maximum likelihood is the most common method of tree-building now

Maximum Likelihood

• idea:

  • find the tree that maximizes the probability of obtaining the observed data

• need:

  • original data

  • an underlying model of evolution

  • trees and their branch lengths

• then:

  • search possible trees and find most likely tree; the tree that maximizes the probability of observing the actual data

Cladogram

• branch length has no meaning

Phylogram

• branch length indicates the amount of evolutionary change

  • corresponds to model of subsitution rate from one nucleotide to another

• ML methods use phylograms to find the tree most consistent with observed sequence data

Statistical Confidence

• we can rarely be certain an entire tree is correct

• instead, we can assign confidence values to parts of the tree

  • how strongly does data support a given clade?

Bootstrap Resampling

• build replicate trees by creating new datasets from original dataset (repeated sampling)

  • pick with replacements

  • create 100s of replicates, and a tree for each

• generate consensus tree indicating hat percentage of the trees each particular branch occurred

• how often do we obtain the same clade?

  • bootstrap values: 0-100% (bad-good)

  • if results are being biased by a few nucleotide sites, branch values will be low

Bootstrap Resampling Results

• 0% (bad) - 100% (good)

  • >70% suggest a particular grouping is reliable

  • 50-70% considered suggestive but not conclusive

  • <50% suggests a particular grouping is not strongly supported by the dataset

Molecular Clocks

• molecular traits change at a steady rate, in a ‘clock-like’ fashion

• recall neutral theory:

  • if mutation rate is constant and generation times are similar,
    # neutral molecular differences between two taxa should be proportional to the age of their most recent common ancestor

• can this be used to date when and how rapidly major events occurred?

  • yes, # substitutions proportional to divergence time

Calibrating Molecular Clocks

• measure genetic distance bewteen two taxa whose divergence is known from:

  • fossil record

  • geological record

• however, we now know that molecular clock runs at different rates in different species

  • even in different regions of the genome within a species

  • rates calibrated for particular gene and lineages might not work for other groups, but can still be useful within a clade

• calibration with fossil record or geological events are only estimates of actual divergence dates

WHIPPO Phylogeny Background

• who are the whales’ closest living relatives?

  • whales, dolphins, and porpoises form a monophyletic group: Cetacea

• relationship between cetaceans and ungulates (horse, hippo, deer, etc.) suggested by skeletal characters

  • proposed as a sister group to artiodactyls (cows, hippos, pigs)

  • outgroup is perissodactyls (horses, rhinos)

• are whales ungulates? or are ungulates paraphyletic?

WHIPPO Synapomorphy

• the astragulus, a bone in the ankle, unites the Artiodactyls

• whales dont have the Artiodactyl-type of astragulus

  • does this imply that whales are outside of artiodactyla?

• on the other hand, whales dont have an astragulus at all!

  • need another kind of data

DNA Synapomorphy WHIPPO

• Beta-casein gene sequencing data suggest that whales are a sister taxa to hippos

• SINE and LINE elements

  • Short of Long INterspersed Elements (transposons)

  • transposition events are rare, and therefore VERY unlikely that 2 homologous SINEs would insert themselves into 2 independent host lineages at exactly the same location

  • reversal is also very unlikely, as these can be detected

  • much more likely to be synapomorphies suggesting that whales are a sister taxa to hippos

Fossil Evidence WHIPPO

• fossils discovered after SINE/LINE analysis have:

  • whalelike characteristic

  • pulley-shaped astragulus

WHIPPO Conclusion

• suggests that traits of hippos and whales that were thought to be convergent adaptations for aquatic life might actually be synapomorphies

• modern taxonomic classification refers to this group as Cetartiodactyla