Assessing what is needed to resolve a molecular phylogeny: simulations and empirical data from emydid turtles
Phylogenies often contain both well-supported and poorly supported nodes. Determining how much additional data might be required to eventually recover most or all nodes with high support is an important pragmatic goal, and simulations have been used to examine this question.
Most simulations have been based on few empirical loci, and suggest that well supported phylogenies can be determined with a very modest amount of data. Here we report the results of an empirical phylogenetic analysis of all 10 genera and 25 of 48 species of the new world pond turtles (family Emydidae) based on one mitochondrial (1070 base pairs) and seven nuclear loci (5961 base pairs), and a more biologically realistic simulation analysis aimed at determining how much more data might be necessary to recover weakly-supported nodes with strong support.
Results: Our mitochondrial-based phylogeny was well resolved, and congruent with some previous mitochondrial results.
For example, all genera, and all species except Pseudemys concinna, P. peninsularis, and Terrapene carolina were monophyletic with strong support from at least one analytical method.
The Emydinae was recovered as monophyletic, but the Deirochelyinae was not. Based on nuclear data, all genera were monophyletic with strong support except Trachemys, and all species except Graptemys pseudogeographica, P.
concinna, T. carolina, and T.
coahuila were monophyletic, generally with strong support. However, the branches subtending most genera were relatively short, and intergeneric relationships within subfamilies were mostly unsupported.Our simulations showed that relatively high bootstrap support values (i.e.
[greater than or equal to] 70) for all nodes were reached in all datasets, but an increase in data did not necessarily equate to an increase in support values. However, simulations based on a single empirical locus reached higher overall levels of support with less data than did the simulations that were based on all seven empirical nuclear loci, and symmetric tree distances were much lower for single versus multiple gene simulation analyses.
Conclusions: Our empirical results provide new insights into the phylogenetics of the Emydidae, but the short branches recovered deep in the tree also indicate the need for additional work on this clade to recover all intergeneric relationships withconfidence and to delimit species for some problematic groups.
Our simulation results suggest that moderate (in the few-to-tens of kb range) amounts of data are necessary to recover most emydid relationships with high support values. They also suggest that previous simulations that do not incorporate among-gene tree topological variance probably greatly underestimate the amount of data needed to recover well-supported phylogenies.
Author: Phillip Q Spinks, Robert C Thomson, Geoff A Lovely and H. Bradley Shaffer
Credits/Source: BMC Evolutionary Biology 2009, 9:56
Published on: 2009-03-12
Copyright by the authors listed above - made available via BioMedCentral (Open Access). Please
make sure to read our disclaimer prior to contacting 7thSpace Interactive. To contact our editors, visit our online helpdesk. If you wish submit your own press release, click here.
Social Bookmarking
RETWEET This! | Digg this! | Post to del.icio.us | Post to Furl | Add to Netscape | Add to Yahoo! | Rojo
|
|
