BEAST estimated phylogeny of anoles. Circles on nodes represent posterior probability, black > 0.95, grey > 0.90, white > 0.70.
In the course of our recent study on sex chromosome evolution in anoles (Gamble et al. in press) [AA post] we assembled a 216-species mitochondrial DNA phylogeny of anoles, the largest published to date (at least that we know of), yet containing only a little more than half of all recognized species. Although we collected new sequences for some species, our dataset is largely built on the hard work of others who collected and published on sequences from across the genus, such as Jackman et al. 1999, Poe 2004, Nicholson et al. 2005, Mahler et al. 2010 [AA post], and Castañeda & de Quieroz 2011 [AA post]. Without access to data from these and other studies, we would have had a far less complete and robust tree for our comparative analyses.
There is a big debate going on now regarding what, where and how much data should be shared in association with publishing academically. I personally feel that providing easy access to those data used and generated during a study serves to accelerate the rate and increase the quality of scientific discovery. I am heartened that more and more journals are making data deposition a requirement for publication, although often this means little more than dumping sequence data to GenBank. Sites like Dryad, Figshare, and GitHub now provide open, permanent, and citable access to raw data, figures and, most importantly in my view, research products like alignments, code and analysis logs. In an effort to make our data as accessible and useful as possible we have archived our alignment, MrBayes and BEAST consensus trees as well as as the BEAST posterior distribution on the digital data repository Dryad [doi link]. It is our hope that other anolologists can use and improve upon these data to ask new, interesting questions and to build a larger, more complete view of the evolution of anoles.
- John David Curlis - March 1, 2022
- Joe Macedonia - January 31, 2022
- Lindsey Swierk - January 27, 2022
Kenneth E. Barnett
Awesome. BEAST- for the phylogeny naive, stands for: “Bayesian Evolutionary Analysis Sampling Trees.” There is a small paragraph at the link below, that explains how the graphics are assembled, based on what parameters etc. :-).
http://evolution.genetics.washington.edu/phylip/software.etc1.html#BEAST
Best,
Kenny
Anthony Geneva
Kenny, thanks for pointing this out, and my apologies to all for not explaining this better. The tree above was estimated using a software package called BEAST, which takes a Bayesian approach to estimate both the relationships among a group of samples, in this case anole species, as well as estimate branch lengths that are correlated with time. Often there are a large number of possible trees that have very similar likelihood. Plotting the posterior probabilities (PP) is a way to summarize support by comparing of a set of trees that all have similar likelihoods and asking how often does a particular relationship appear in that set. If PP=1 then all of those trees contain that node if it’s 0.5 only have of the probable set of trees have that relationship.
Jonathan Losos
This is the latest, most up-to-date, anole phylogeny. Those of you keeping score at home will recall that the previous most up-to-date phylogeny agreed almost completely with the eight genera-level clades suggested by Nicholson et al. How do the N et al. genera fare in this phylogeny? Not quite as well. Here are the problems: A. koopmani falls within the Deiropteryx clade rather than the Chamaelinorops clade. The clade comprised of A. argenteolus + A. lucius falls in the Xiphosurus clade rather than the Anolis clade; A. christophei falls in the Xiphosurus clade rather than the Chamaelinorops clade. Hence, unless I overlooked other disagreements, five of the N et al. clades are found in this new tree, but the other three are not monophyletic.
Nico Franz
To reconcile multiple, recent Anole phylogenies and classifications, would you consider exploring this “taxonomic concept alignment” prototype tool? http://taxonbytes.org/using-the-euler-x-toolkit-to-align-taxonomies-introductory-notes/
This is particularly useful for mapping alternative, succeeding trees, using logic representations and reasoning. A small example use case is shown here: http://taxonbytes.org/alignment-of-two-classifications-of-the-monotremata/
Jonathan Losos
I realize now that the previous most recent phylogeny, by Pyron et al., also disagreed with Nicholson et al. in the placement of A. christophei: Nicholson et al. put it in their Chamaelinorops clade, but the Pyron et al., tree, like the Gamble et al. tree, place it with the Nicholson et al. Xiphosurus species. This was an oversight of mine in my previous post; I’m not sure how I made that mistake, but in checking, I realize that Nicholson et al. refer to that species as both Chamaelinorops christophei and Xiphosurus christophei in their paper, which may have been the source of my error.
Liam Revell
Cool.
How was the plot created? You have something weird going on with your node labels. (They are not centered on the nodes & move around depending on their position in the tree.)
Here is a plot created using plotTree in phytools and nodelabels in ape from the BEAST tree file on Dryad:
Liam Revell
My plot didn’t seem to post for some reason. Here’s a second attempt:
Liam Revell
One more try:
Anthony Geneva
Thanks Liam Your version does look better, although I’m partial to colored circles, versus pie charts for support. I can’t remember how I originally made that figure (and can’t find the script I used), it actually from over a year ago. I typically use plot.phylo() and nodelabels() in ape.