I have a dataset concerning mtDNA haplotypes with restriction
site data, and the frequency of these haplotypes across 12
populations. I want to obtain a dendrogram with support
levels at the nodes that displays some genetic distance
between populations. I can bootstrap the haplotype data to
show relatedness of haplotypes but want to do a similar thing
with populations. I have the restriction enzyme analysis
package (REAP, McElroy et al.) and can get a single measure
of nucleotide divergence between populations but some
populations have -ve values due to high diversity. After
correcting to zero the UPGMA dendrogram of this clustered
data yields an equally likely number of trees and the
concensus invloves an 11 population unresolved node. I really
want a more informative tree. Do I have to feed the
bootstrapped haplotype datasets back into REAP to produce 100
(possibly) different population trees and put these into
consense in say PHYLIP, or is there an easier way?
Craig Wilding
Leeds University
