[Taxacom] Phylogenetic networks

[Bob Mesibov]

A 'phylogenetic network' nowadays means a messy web of possible relationships that tree-building software for phylogenetic analysis can't handle. IMHO this failure has led some workers to have a blind spot with regard to the possibility that their carefully built trees are the result of squeezing character data into a mold they don't fit.

But software does, in fact, exist for building networks from molecular character data, and what you get out is a kind of network overlay which shows different relationship possibilities all at once, including some your Bayesian, ML and parsimony methods might have said were well supported. (You don't have to say 'This is the best supported set of clades' and publish a tree. You can also say 'These are the best supported sets of relationships' and publish a network. Both approaches clearly show groups.)

Network analytical software has been widely used in recent years for exploring relationships between closely related exemplars, and less frequently for looking at cases of horizontal gene transfer. There is now a textbook on the topic: http://www.phylogenetic-networks.org/ , and at the risk of being accused of using TAXACOM for advertising, I draw attention to it here. I have no connection with the authors or the publisher, but I've used SplitsTree three times: twice to explore arachnid clades published by cladists as trees, and once to double-check a colleague's networks of intraspecific 16S and COI relationships (paper submitted).
-- 
Dr Robert Mesibov
Honorary Research Associate
Queen Victoria Museum and Art Gallery, and
School of Zoology, University of Tasmania
Home contact: PO Box 101, Penguin, Tasmania, Australia 7316
Ph: (03) 64371195; 61 3 64371195
Webpage: http://www.qvmag.tas.gov.au/?articleID=570