Re: Long branch attraction

[Phillip Bigelow <bigelowp@juno.com>]

Roberto Takata <rmtakata@gmail.com> writes:
> As longer a branch is, more mutations are inferred to occurred in the 
> lineage.


So, if there are only 2 steps between the two basal-most taxa on a
branch, but there are 15 steps between them and the next taxa "higher up"
on the branch, would that be an example of a "long branch"?  And should
this topology be held suspect?


> 
> More mutation there are, more probable that homoplasies arise.
> Since
> we assume maximum parcimony, homoplasies are prone to be taken as
> homologies...


So, the problem with having many steps between taxa is that it increases
the risk of the computer algorithm misinterpreting homoplasies as
homologies.  Therefore, distantly-related taxa get "attracted" to one
evolutionary branch, when they actually should be on put on different
branches.

How'm I doin'?

<pb>
--
"If you are traveling into the future in a time machine, and you pass a
person from the future who is traveling into the past, it's probably a
good idea to avoid eye contact." - Jack Handey