7. Rooting & Polarity Estimation

Review this lecture; a lot of it is visual.

Long Branch Attraction (LBA) occurs when…

When there is a

• High frequency of parallel, or

• Convergent changes causing sequences (or other characters) to arrive at the same character state

After extended periods or accelerated rates of evolution.

Define ‘Root‘

The earliest node in a tree, the point from which the descendent lineages arose

How to root a tree: Outgroups

Rooting of trees relates to polarization of character states.

The use of outgroups roots the tree and helps polarize character change.

Ex) Without outgroups: plesiomorphic state could be either A or T.

Multiple outgroups are preferable: Use of multiple outgroups can help resolve these uncertainties, especially if they exhibit the same state for the relevant character.

How to root a tree: The Fossil Record

• Can also be used to polarize character change.

• Other things being equal, the condition appearing earlier in the fossil record is plesiomorphic.

• The sequence of fossil appearance is based on superposition of rock layers.

• But: the fossil record is incomplete, and many characters are not preserved in fossils.

How to root a tree: Developmental (embryological) Sequence

• Has also been used to polarize character change.

• The now-disregarded ”biogenetic law” assumes that the condition appearing earlier in development is plesiomorphic.

• “von Baer’s law” is better supported by evidence

How to root a tree: Midpoint Rooting

This method of rooting can be applied to datasets with large numbers of characters, such as DNA sequence data.

Places the root halfway between the two most distant taxa.

Relies on the assumption that evolution is proportional to time (divergence rates are constant).

How to root a tree: Gene-duplication Rooting

This method of rooting can be applied to certain classes of molecular sequence data. It is very compelling when it can be used.

Orthologs are different forms of a single gene that have evolved by speciation within a study group. Functions usually remain similar.

Paralogs are the alternative forms of a duplicated gene that evolve somewhat independently within a study group after the duplication event. Functions usually diverge among paralogs, given time.

Homologs can be either orthologs or paralogs.