Hi all,
I am new to phylogeny, please excuse me if this question is trivial or if I make some conceptual mistakes.
I need to build a consensus tree of a collection of aka 40 orthologs of a protein that has a globular domain and an intrinsically disordered region. I already have the multiple sequence alignment. Essentially, I have no time or computer power limitations.
Currently, I am using this very simple MrBayes script:
begin mrbayes;
set autoclose=yes nowarn=yes;
[data file]
execute myalignment.nex;
[substitution model]
lset nst=1 rates=gamma;
[mcmc settings]
mcmc savebrlens=yes nruns=1 ngen=100000 samplefreq=10;
sumt burnin=1000;
end;
Do you guys have any recommendation on how to fine-tune this script in order to better fit my needs?
Any information will be helpful!
Thanks in advance,
Miquel
•
written
by
miquelduranfrigola
Hi Miquel,
I think nobody can answer your question as it is now. I can offer some advice though. The software and parameters you use will entirely depend on the question you want to answer, what is it? You want to reconstruct the ancestral state, infer a species phylogeny... Depending on your question you will select one optimality criterion or another that will lead you to a specific method and implemented by different softwares. I would recommend a classical book: Infering phylogenis by J. Felstenstein. Good starting point.
Cheers
Hi Biojl, thanks for the recommendation. I'll have a look and re-formulate my question in the next days.
Hi Miquel,
The first step in any ML or Bayesian approach is to define the model of evolution that best fits your data.
If your alignment is DNA you can use "jmodeltest". For proteins, you can use "ProtTest". It would be good to read the manuals (and associated papers) to these pieces of software as they explain how the model is selected.
Given that model, read the MrBayes manual on how to implement the model. Usually you will want to edit the "lset" line in your script.
Hope this helps!
Cool. I definetely miss the ProTest step.
Log in to answer this question.
