TPM2uf+G model in BEAUTI

[Michael Jowers]

Dear all,

Would any one know how to change the Operators (or what the XML code is) to run a TPM2uf+G model in BEAUTI?

details are as follow below:

Many thanks!

Michael

TPM2uf (nst=6, AC=AT CG=GT, AG=CT)

    partition = 010212

   -lnL = 733.3953

   K = 128

   freqA = 0.3496

   freqC = 0.2145

   freqG = 0.1281

   freqT = 0.3078

   R(a) [AC] =  0.0647

   R(b) [AG] =  2.8708

   R(c) [AT] =  0.0647

   R(d) [CG] =  1.0000

   R(e) [CT] =  2.8708

   R(f) [GT] =  1.0000

   gamma shape = 0.0230

[Remco Bouckaert]

Hi Michael,

In BEAUti, you can do this as follows: Select GTR as substitution model. Then, select the menu Mode/Allow parameter linking and a link icon appears next to the rates of GTR. If you click the icon, a list of parameters pops up that you can link with. To create a TPM2 model, you just link AC with AT, CG with GT and AG with CT. By default, CT is fixed at 1, so linking AG with CT causes AG to be fixed to 1 as well. If you would estimate them they would not be identifiable with tree height and clock rate, so you want at least one of the parameters to be fixed.

For the frequencies, make sure to keep them a ‘estimated’ and set Gamma category count to 4, and select the ‘estimate’ box next to the shape paramater to completely specify TPM2uf+G.

If you are not interested in the parameters of this particular model, but just want to specify a suitable site model, you might want to consider the BEAST model test site model, which is available if you install the bModelTest package (paper in print in BMC Evolutionary Biology). It does Bayesian model averaging over a set of site models and TPM2ug+G is one of the models that is averaged over.

Cheers,

Remco

--
You received this message because you are subscribed to the Google Groups "beast-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to beast-users...@googlegroups.com.
To post to this group, send email to beast...@googlegroups.com.
Visit this group at https://groups.google.com/group/beast-users.
For more options, visit https://groups.google.com/d/optout.

[Michael Jowers]

Dear Remco,

Many thanks for the detailed explanation. Can this be done similarly in Beast 1 or is it only possible in Beast 2?

Best,

Michael

To unsubscribe from this group and stop receiving emails from it, send an email to beast-users+unsubscribe@googlegroups.com.

To post to this group, send email to beast...@googlegroups.com.
Visit this group at https://groups.google.com/group/beast-users.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "beast-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to beast-users+unsubscribe@googlegroups.com.

[Remco Bouckaert]

Hi Michael,

The description is for BEAUti 2 — I’m not sure whether it is possible to do this in BEAUti 1, though you certainly can do it by editing the XML (e.g. see http://beast.bio.ed.ac.uk/Nucleotide-Substitution-Models).

bModelTest is a package for BEAST 2 only.

Cheers,

Remco

To unsubscribe from this group and stop receiving emails from it, send an email to beast-users...@googlegroups.com.