Patrick,
I am working on these kinds of issues as well.
If you use Arlequin, or Hapstar to make networks, you will not get loops 
either, at least when you use the primary haplotype.  TCS does keep loops, and 
separates networks without connections.

Giacomo


On Mar 29, 2011, at 7:28 AM, Patrick Flight wrote:

> Hello r-sig-phylo,
> 
> I am using the haploNet command of the excellent pegas package to construct 
> an mtDNA haplotype network with 79 nodes. When I generate the network I am 
> able to identify the limits of parsimony from the accompanying table. 
> However, I don't quite understand why this network contains no loops (79 
> nodes 78 edges). When I make the network in TCS the parsimony limits are the 
> same (7 steps), but there are many loops in the network. The documentation 
> refers to the Templeton et al. 1992 paper and, short of identifying 
> recombination to break loops, I couldn't find hard and fast rules for 
> breaking loops (I may have missed them). I know that some post hoc rule 
> exist, such as favoring links to high frequency and geographically close 
> haplotypes, but I was wondering how haploNet deals with these loops. Does it 
> present 1 of the maximum parsimony trees?
> 
> Thanks in advance,
> Patrick
> 
> _______________________________________________
> R-sig-phylo mailing list
> [email protected]
> https://stat.ethz.ch/mailman/listinfo/r-sig-phylo

Giacomo Bernardi
Professor, Department of Ecology and Evolutionary Biology
University of California
Center for Ocean Health, Long Marine Lab
100 Shaffer Road
Santa Cruz, CA 95060, USA

Tel: office 831 459 5124
Tel: lab 831 459 1282
Fax: 831 459 3383
email: [email protected]
http://bio.research.ucsc.edu/people/bernardi

_______________________________________________
R-sig-phylo mailing list
[email protected]
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo

Reply via email to