Dear colleagues;
I am approaching Bayiesian phylogenetic analysis, and I am dealing with (fossil) morphological datasets. I am especially interested in tip-dating and I would like to have an opinion which software is best between BEAST and Mr.Bayes, considering also the presence of multi-state characters in the matrix.
Thank you very much in advance.
Filippo
I am fairly sure BEAST is only used for molecular data so if you have other types you would be using MrBayes (or RevBayes?)
Hi,
I recommend MrBayes. Try to run it via the CIPRES portal which will save you a lot of computing time .
Best,
Sophie
Thank you very much for your suggestions !
Rutger A. Vos , BEAST has been used also for morphological data in some recent works (see, for instance, this paper: https://science.sciencemag.org/content/345/6196/562 )
Best to all
Filippo
Warning about Bayesian phylogenetics with paleontological datasets (i.e. plenty of missing data that are distributed neither randomly nor by rate of evolution): we don't know if it works. There has never been a systematic test. All we have are a few empirical examples... and they suggest to me that Bayesian inference takes some of the mistakes parsimony makes, and then doubles and triples down on them till they get posterior probabilities > 75%. There's a short discussion, with plenty of references, in PeerJ paper 5565, I think on page 96.
David Marjanović thank you for your insight. Can you be more precise regarding the exact reference you're talking about?
Sorry. It's https://peerj.com/articles/5565 , on pp. 96–99 of the PDF. Tell me if you have further questions!
David Marjanović thank you very much! I am reading it and I find it very interesting. Do you mind if I email you if I have more questions?
- Badges
- Science topic