BiSSE/MacroCAIC on Non-Ultrametric Tree with Polytomies?
Hi again...brief thought -- I am so used to thinking about diversification in terms of speciation and extinction rates in units of time, it's hard to think about how a study of correlation between a trait and diversification/species richness would work without a time-scaled tree. However, there are some methods that use tree-balance statistics, and thus only require topology -- see work by Brian Moore. I think they require pretty dramatic differences between sister groups to infer rate shifts though. Cheers! Nick
On Fri, Apr 15, 2016 at 6:59 AM, Brian A. Gill <gillbriana at gmail.com> wrote:
Hi David and Joe. Thanks for your input/questions. David: My tree is non-ultrametric because it is undated tree of extant taxa. It does not include taxa from the fossil record. Making a dated tree seems like a good option, but what about the option of using NonParametric Rate Smoothing (NPRS) or some other conversion to make my existing tree ultrametric? I see this approach in some studies. Alternatively, sister-clade comparisons could be an option, but it seems like they are old and not powerful. Based on my tree topology and the number of taxa, I'm not sure I have enough contrasts for a robust test. Joe: For BiSSE, the discrete binary predictor is assumed to evolve according to a Markov model along the tree. The model is then used to estimate speciation, extinction, and character transition rates. For MacroCAIC, I assume the a discrete binary trait would be modeled in a similar way to determine the states of internal nodes on the tree. So, thinking about evolution of a discrete trait by a Markov process on a non-ultrametric tree, it seems for BiSSE having non-contemporaneous lineages would inappropriately inflate the extinction parameter. For MacroCAIC, I can't think of a reason why a non-ultrametric tree would be problematic other than how differences in lengths of branches based on relative divergence vs. absolute time might affect estimates of states at internal nodes. If this is the case, does that mean that other procedures using Markov transition matrices cannot be used on trees with relative divergence for branch lengths (e.g. stochastic character mapping)? Thanks for all the help. Brian On Thu, Apr 14, 2016 at 8:26 AM, David Bapst <dwbapst at gmail.com> wrote:
Brian- Is your tree non-ultrametric because it contains extinct taxa from the fossil record, or is it simply an undated tree of extant taxa? To my knowledge, there isn't yet a satisfactory solution to the first (no BISSE for paleo-phylogenies) but if the second then, it seems the best route would be to date it and use the methods developed for ultrametric trees. Cheers, -Dave On Wed, Apr 13, 2016 at 3:09 PM, Brian A. Gill <gillbriana at gmail.com> wrote:
Hi Everyone. I'm trying to look at the association between a discrete binary
predictor
(Latitude: Colorado/Ecuador) and a continuous response (species
richness).
My phylogeny is a 50% majority rule consensus tree made in MrBayes. The tree has polytomies and is not ultrametric. I've found the methods below for looking at the influence of a discrete binary trait on richness, but I'm not sure if my tree is suitable or if there is a better approach. 1) Diversitree package BiSSE 2) Caper package using MacroCAIC Any suggestions would be greatly appreciated. Thank You. Brian -- Brian A. Gill VISIT MY WEBSITE: http://gillbriana.wix.com/brian-gill FOLLOW ME ON TWITTER: @CSUBrianGill Colorado State University Biology 1878 Campus Delivery Fort Collins, CO 80523 United States of America 970-215-7037 [[alternative HTML version deleted]]
_______________________________________________ R-sig-phylo mailing list - R-sig-phylo at r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo Searchable archive at
http://www.mail-archive.com/r-sig-phylo at r-project.org/ -- David W. Bapst, PhD Adjunct Asst. Professor, Geology and Geol. Eng. South Dakota School of Mines and Technology 501 E. St. Joseph Rapid City, SD 57701 http://webpages.sdsmt.edu/~dbapst/ http://cran.r-project.org/web/packages/paleotree/index.html
-- Brian A. Gill VISIT MY WEBSITE: http://gillbriana.wix.com/brian-gill FOLLOW ME ON TWITTER: @CSUBrianGill Colorado State University Biology 1878 Campus Delivery Fort Collins, CO 80523 United States of America 970-215-7037 [[alternative HTML version deleted]]
_______________________________________________ R-sig-phylo mailing list - R-sig-phylo at r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo Searchable archive at http://www.mail-archive.com/r-sig-phylo at r-project.org/