computing pairwise ts/tv.

Sudhir Kumar IMEG at psuvm.psu.edu
Tue Aug 16 21:24:53 EST 1994


Much of the discussion about transition/transversion ratios in this forum has
been directed towards nucleotide sequence data that has more than 2 sequences.
 To get an idea regarding ts/tv bias, we can also compute this ratio for every
pair of sequences in the data.  In computer program MEGA, all pairwise ts/tv's
can be computed in one run along with their standard errors.  In addition, one
can also compute distances based only on transversional substitutions, if one
suspects that transitional substitutions are close to saturation.  This can be
done for most of the distance estimation methods available, such as the ones
that do not require G+C-content to be equal to 50% and the ones that require
more than 3 parameters for modelling the substitution patterns.
                                                   Sudhir Kumar



More information about the Mol-evol mailing list