[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Resetting the "molecular clock hypothesis"

> However, the mastadon/mammoth case makes this rather
> irrelevant - it seems we are talking about convergent
> evolution on the molecular scale, which I haven't heard
> of before,

You never heard about long branch attraction in moecular phylogenetics? ;-)

(OK, I know that's not what you mean. But simple statistics will show molecular 
convergence on the *point mutation* level pops up quite quickly in some 
sequences. Consider a ~2%/Ma mutation rate for cytB commonly found in rodents 
and passeriforms; statistically every base in that sequence has mutated once 
between the most divergent lineages of these two, and since mutations are not 
equally distributed you can expect *some* amount of convergence here to pop up).

Then we have the case of large-scale events, though it may be hard to 
distinguish convergence from retained plesiomorphisms. See for example here: 
"...nucleotide insertionâdeletions that support mutually exclusive 
phylogenetic hypotheses in different gene trees." (they conclude that it's 
retained plesiomorphies inconsistently retained through incomplete lineage 
sorting). In active retroposon insertion hotspots on the other hand it is just 
as likely to be convergence.

It is not common enough, but this is of little comfort when the sequence you 
want to use has undergone such "not common" events. And the main drawback of 
molecular data - homoplasy is probably more common in anatomy, but if the scope 
of your study is not overly large (say Livezey/Zusi-large), you can tailor your 
characters so that it is not that much of a problem. For sequence data where 
there are just 4 possible character states and with each change of state the 
previous state gets entirely obliterated, it is more problematic. So you can 
prevent the "Gaviomorphae" from popping up by tricks such as not treating 
cnemial crests as a binary-state (present/absent) character. But you cannot 
*know* what some base position has changed from, only what it has changed to 
here and now.



Gesendet von Yahoo! Mail.
Dem pfiffigeren Posteingang.