Molecular Biology and Evolution, Vol 15, 727-737, Copyright © 1998 by Society for Molecular Biology and Evolution
N Takezaki
In examining genetic data in recent publications, Backeljau et al. showed
cases in which two or more different trees (tie trees) were constructed
from a single data set for the neighbor-joining (NJ) method and the
unweighted pair group method with arithmetic mean (UPGMA). However, it is
still unclear how often and under what conditions tie trees are generated.
Therefore, I examined these problems by computer simulation. Examination of
cases in which tie trees occur shows that tie trees can appear when no
substitutions occur along some interior branch(es) on a tree. However, even
when some substitutions occur along interior branches, tie trees can appear
by chance if parallel or backward substitutions occur at some sites. The
simulation results showed that tie trees occur relatively frequently for
sequences with low divergence levels or with small numbers of sites. For
such data, UPGMA sometimes produced tie trees quite frequently, whereas tie
trees for the NJ method were generally rare. In the simulation, bootstrap
values for clusters (tie clusters) that differed among tie trees were
mostly low (< 60%). With a small probability, relatively high bootstrap
values (at most 70%-80%) appeared for tie clusters. The bias of the
bootstrap values caused by an input order of sequence can be avoided if one
of the different paths in the cycles of making an NJ or UPGMA tree is
chosen at random in each bootstrap replication.
ORIGINAL ARTICLE
Tie trees generated by distance methods of phylogenetic reconstruction
National Institute of Genetics, Shizuoka, Japan. ntakezak@lab.nig.ac.jp
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. Meier, K. Shiyang, G. Vaidya, and P. K. L. Ng DNA Barcoding and Taxonomy in Diptera: A Tale of High Intraspecific Variability and Low Identification Success Syst Biol, October 1, 2006; 55(5): 715 - 728. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Arnau, S. Mars, and I. Marín Iterative Cluster Analysis of Protein Interaction Data Bioinformatics, February 1, 2005; 21(3): 364 - 378. [Abstract] [Full Text] [PDF]
|
|