comparing results with different -pm of DPP priors and -urg prior

[Xin Chen]

Hi,

For my dataset, I believe the rate for this tree won't fit strict clock, based on a test of all gene trees, but I am not sure what could be the proper value of -pm or should I use -urg. Thus, I setup three runs with 1. default -pm =3; 2. -pm=20; 3. using -urg (my tree has 87 taxa). May I ask if it is possible to compare results from multiple runs? 

Thank you so much, 

Xin

[Tracy Heath]

Hi Xin Chen,

The -pm is the prior mean number of rate categories, you should assign this a value that is consistent with your prior belief of the number of different branch rates. So since I am unfamiliar with your data, I cannot recommend a value for you to specify. However, by default, DPPDiv places a hyperprior on the concentration parameter of the DPP, thus you can account for your uncertainty in this parameter. If the data are informative on branch rates, then you will find that, with this hyperprior, the posterior number of categories is robust to the prior mean (see figure 8 in our paper: http://phylo.bio.ku.edu/sites/default/files/Heath_etal_MBE_2012.pdf). If you compare the results from your different runs (1 and 2), then you can do this to assess the sensitivity of your posterior to the priors. Though you should do this after you decide which prior you will use for the run from which you will summarize results. 

Cheers,

Tracy

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

[Luis]

Hi Tracy,

How are you?

I am running the last version of dppdiv. I'm very happy because I could finally compile it :)

Yet, I could not find the option for running my data under the GTR+Γ model of sequence evolution or the relaxed-clock model option in order to allow substitution rates to vary across the tree. Can you help me with this?

Cheers

Luis

[Tracy Heath]

Hi Luis,

It's nice to hear from you. 

The only sequence model available in DPPDiv is GTR+G, so there is no setting for this option. Also, by default the program uses a DPP prior on the branch rates (see http://phylo.bio.ku.edu/sites/default/files/Heath_etal_MBE_2012.pdf). You can alter the hyperprior on the number of branch-rate categories by using the "-pm" option. 

Most of the options are listed in the help menu. If you run DPPDiv without any arguments or with the "-h" argument, then it will provide the list for you.

Let me know if you have further questions.

Cheers!

Tracy