Just to add to Brian and Aaron’s comments, you also want to consider what those alpha values mean in relation to your data. The alpha’s you mention equate to phylogenetic half-lives of between 0.17 and 0.28 time units, which are extremely rapid rates of adaptation if your branch lengths are in millions of years and your clade is old--ish (i.e. 10s of millions of years). But they’re perhaps less dramatic if you’ve rescaled the root - tip distance of the tree to 1 prior to analysis, or if your clade is just a couple of million years old. To be sure, checking the likelihood surfaces and your ability to identify the parameters for a given dataset is a crucial step in any analysis. But those numbers are essentially meaningless without being placed in the context of your data.
Graham ------------------------------------------------------------ Graham Slater Peter Buck Post-Doctoral Fellow Department of Paleobiology National Museum of Natural History The Smithsonian Institution [NHB, MRC 121] P.O. Box 37012 (202) 633-1316 slat...@si.edu<mailto:slat...@si.edu> www.fourdimensionalbiology.com<http://www.fourdimensionalbiology.com> On Mar 27, 2015, at 2:58 PM, Brian O'Meara <omeara.br...@gmail.com<mailto:omeara.br...@gmail.com>> wrote: You can select between models using AICc, but then you can look at the reliability, more robustly than using eigenvalues, by using OUwie.boot() to do parametric bootstrapping. Another very useful thing done by Hansen, Bartoszek, and colleagues is to do a contour plot around the maximum likelihood estimate to see how much a parameter like alpha can vary without changing the likelihood much. As Aaron indicated, this can be a horrifyingly large region. It's especially tough with OUMA and OUMVA where there are multiple alphas. One thing to consider is what kind of model is most likely to give you results you trust and that are relevant for your question. It could be that you care about OUM-type models, but don't really care if it's OUMA, OUM, or OUMVA. You should certainly report that, say, OUMVA is the best fitting under AICc, but perhaps if your biological question is about the optima only you focus on OUM due to OUMVA seeming to have issues estimating any parameters well (though of course also indicate if the results under OUMVA generally agree or disagree with the conclusions from OUM -- you need to show that it's not that OUMVA fit better but gave you an answer you didn't like, so you cherry picked a different model to get the result you want). Hope this helps, Brian _______________________________________ Brian O'Meara Assistant Professor Dept. of Ecology & Evolutionary Biology U. of Tennessee, Knoxville http://www.brianomeara.info Postdoc collaborators wanted: http://nimbios.org/postdocs/ Calendar: http://www.brianomeara.info/calendars/omeara On Fri, Mar 27, 2015 at 12:57 PM, Diego Salazar Tortosa <dsala...@ugr.es> wrote: Thank you for your answer Aaron. Clear to me that this situation is common in the OU models, but I don't know what criteria I should use when select between different OU models, AICc or reliability of parameters? Thanks in advance Regards Diego Salazar 2015-03-27 11:08 GMT+00:00 Aaron King <kin...@umich.edu>: In many situations, the OU model parameter alpha is not well identified. This has been pointed out before, but it's quite common for the parameter values (and not just alpha, though that's typically the worst) to be poorly identified, even if the model selection is unambiguous. OU model parameters can be well identified under some circumstances (detailed in the paper) but when they are not, estimates can be *very* unreliable. On Fri, Mar 27, 2015 at 4:07 AM, Diego Salazar Tortosa <dsala...@ugr.es> wrote: Hi, I am trying to analyze whether the evolution of a continuous trait is state-dependent of discrete trait with three levels, through OU models. My phylogeny has 113 species and 112 internal nodes. Clearly the most parsimonious models according AICc are OUMA and OUMVA but have alpha values between 2.5 and 4, are these values usual or too high? Also some eigenvuales obtained with diagn=T are negative for both models, althought standard errors are very small. Are these models valid? Thanks in advance Diego Salazar -- Diego Francisco Salazar Tortosa Ph student Departamento de Ecología Facultad de Ciencias Universidad de Granada Av. Fuente Nueva s/n 18071 Granada Telefono: +34 958241000 ext 20007 Movil: +34 634851132 email: dsala...@ugr.es <die...@correo.ugr.es> dftort...@gmail.com ------------------------------------------------------------------- "Este mensaje se dirige exclusivamente a su destinatario y puede contener información privilegiada o confidencial. Si no es Ud. el destinatario indicado, queda notificado de que la utilización, divulgación o copia sin autorización está prohibida en virtud de la legislación vigente. Si ha recibido este mensaje por error, se ruega lo comunique inmediatamente por esta misma vía y proceda a su destrucción. This message is intended exclusively for its addressee and may contain information that is CONFIDENTIAL and protected by professional privilege. If you are not the intended recipient you are hereby notified that any dissemination, copy or disclosure of this communication is strictly prohibited by law. If this message has been received in error, please immediately notify us via e-mail and delete it". ---------------------------------------------------------------------- [[alternative HTML version deleted]] _______________________________________________ R-sig-phylo mailing list - R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo Searchable archive at http://www.mail-archive.com/r-sig-phylo@r-project.org/ -- Aaron A. King, Ph.D. Ecology & Evolutionary Biology Mathematics Center for the Study of Complex Systems University of Michigan GPG Public Key: 0x15780975 -- Diego Francisco Salazar Tortosa Ph student Department of Ecology University of Granada Av. Fuente Nueva s/n 18071 Granada Telefono: +34 958241000 ext 20007 Movil: +34 634851132 email: dsala...@ugr.es <die...@correo.ugr.es> dftort...@gmail.com ------------------------------------------------------------------- "Este mensaje se dirige exclusivamente a su destinatario y puede contener información privilegiada o confidencial. Si no es Ud. el destinatario indicado, queda notificado de que la utilización, divulgación o copia sin autorización está prohibida en virtud de la legislación vigente. Si ha recibido este mensaje por error, se ruega lo comunique inmediatamente por esta misma vía y proceda a su destrucción. This message is intended exclusively for its addressee and may contain information that is CONFIDENTIAL and protected by professional privilege. If you are not the intended recipient you are hereby notified that any dissemination, copy or disclosure of this communication is strictly prohibited by law. If this message has been received in error, please immediately notify us via e-mail and delete it". ---------------------------------------------------------------------- [[alternative HTML version deleted]] _______________________________________________ R-sig-phylo mailing list - R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo Searchable archive at http://www.mail-archive.com/r-sig-phylo@r-project.org/ [[alternative HTML version deleted]] _______________________________________________ R-sig-phylo mailing list - R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo Searchable archive at http://www.mail-archive.com/r-sig-phylo@r-project.org/ [[alternative HTML version deleted]] _______________________________________________ R-sig-phylo mailing list - R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo Searchable archive at http://www.mail-archive.com/r-sig-phylo@r-project.org/
