[Mesquitelist] simulation issue: state frequencies and scaling factor

[Paul Bacquet]

Dear all,

I have two questions concerning the use of mesquite simulation tool for 
character evolution.

I would like to simulate a sequence of binary characters on a tree to see if the observed diversity in real data is stronger than expected by chance.
Firstly I need to constrain the state frequencies on this sequence but I can't find how to do this with Markov models implemented in Mesquite.
The solution I found is to use the DNA simulation tool, allowing only the substitution between two states (example A<->G) and using a "user specified nucleotide frequencies" in the GTR options.
It seems to work but maybe is there a more conventional way to do this?
Secondly I don't know which scaling factor value I should use.
I first have estimated the rate of evolution with my real data on the species tree with BayesTraits program (Multistate) and obtained a value of 60.
Should I use a scaling factor of 60? (I tried but then the phylogenetic signal observed in my data wasn't anymore present in the simulated character matrices so I guess the value was inappropriate)

I hope you can help me on these two points

best regards

Paul

PhD student
Evolutionary Ecology and Genetics
Université catholique de Louvain
Belgium


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mesquiteproject.org/pipermail/mesquitelist/attachments/20110925/a2f6a90d/attachment.html