Steve Hardies asks (about my suggestion to jointly maximize
transition/transversion ratios and the tree in a maximum likelihood method):
> Do you think it would be useful (when using the above method) to obtain
> the transition/transversion ratio from a subset of the most closely
> related sequences to suppress noise caused by saturation of
> transitions at greater divergence?
No, as ML will automatically correct for multiple overlaying substitutions
and for the greater uncertainty as to what the signal from distance sequences
means. Thus (given the correctness of the model it uses, of course) it
will make optimal use of the information from closely related and distantly
related sequences.
-----
Joe Felsenstein, Dept. of Genetics, Univ. of Washington, Seattle, WA 98195
Internet: joe at genetics.washington.edu (IP No. 128.95.12.41)
