Marine lab researchers are back, and network speeds have just about given up (down to bytes/s). This post is composed offline and will be uploaded to the notebook. Details are well logged on the Bodgea Phylogenetics Wiki so these are reasonably sketchy with my comments.
My Updates to Bodega Phylogenetics Wiki
Made several contributions to the Bodega Phylogenetics Wiki today:
- User page
- Phylogenetics on Linux
- Becoming a Programmer
- Edits on Advice for Aspiring Phylogeneticists
Today’s agenda
- BEAST continued. (Brian Moore)
- Species Tree and Gene Trees (Bob Thompson)
- RAxML Tutorial (Greg Pauly)
- Model selection and MCMC diagnosis (Brian Moore)
- R tutorial (Rich Glor)
Brian Moore: BEAST
8:30 - 9:30a Diving into the XML and actually running BEAST. Wiki page.
Bob Thompson: Species Tree vs Gene Tree
9:45
- Coalescent! Why don’t species tree and gene tree agree?
- Incomplete Coalescence (Typically we focus on this.)
- Gene duplication
- Horizontal transfer
Differences are common and Widespread! Pollard et. al. (2006), PMID:17132051
BEST: Bayesian Estimation of Species Trees Works with MrBayes, build on MrBayes
Greg Pauly, RAxML
Really large trees can be done no other way. (Garli works, but doesn’t permit partitions or amino acid based inference).
Rich Glor, Introduction to R
1:30 p
“We’re going into combat, but we’re not the Marines. The Marines have this slogan Leave no man behind. This isn’t like that.” -R. Glor
Brian Moore, Model Selection and MCMC diagnosis
slides Estimation under the prior alone!
Brian’s scripts for keeping track of what parameters he’s tried. Very interesting idea.
(We didn’t actually get through all 181 slides…)