BEAST2 packages on Linux

[Manon Ragonnet]

Hi there,

I am running BEAST2 v2.4.0 on a Linux cluster. I am trying to run an xml which makes use of the BDSKY and BEASTLabs packages, which I have installed. If I type addonmanager -list, those packages display as installed.

However, the xml will not run, it crashes with an exit error indicating that beast2 cannot access the packages on the server.

"Error 1017 parsing the xml input file

nullClass could not be found. Did you mean beast.evolution.speciation.BirthDeathGernhard08Model?

Error detected about here:

  <beast>

      <run id='mcmc' spec='MCMC'>

          <distribution id='posterior' spec='util.CompoundDistribution'>

              <distribution id='prior' spec='util.CompoundDistribution'>

                  <distribution id='BirthDeathSkySerial.t:AC_1,2' spec='beast.evolution.speciation.BirthDeathSkylineModel'>"

How can I make sure beast2 access the packages?

The xml runs on my desktop, and when Denise Kuhnert provided me with a jar file in which the package is wrapped, it runs as well - but this isn't a good solution for me as I want to use the other packages as well.

Thanks,

Manon

[Huw A. Ogilvie]

Does beast2 run without errors on the head node of the cluster? If so, it would be best to work with your cluster administrators to fix their/your configuration.

[Manon Ragonnet]

Edit: it's a Linux server not a cluster. BEAST2 runs no problem as long as the xml doesn't require any packages.

[Remco Bouckaert]

Hi Manon,

To install packages on a server, you can use the addonmanager from the commend line.

If you installed beast in say /path/to/beast, then the addonmanager is in /path/to/beast/bin/ 

To see which packages are installed, use 

/path/to/beast/addonmanager -list

To install  a package, say BDSKY, use

/path/to/beast/addonmanager -add BDSKY

which will be installed in your home directory in ~/.beast/2.4/BDSKY.

To remove  a package, say BDSKY, use

/path/to/beast/addonmanager -del BDSKY

There is more information here: http://beast2.org/managing-packages

Cheers,

Remco

On 5/05/2016, at 9:47 PM, Manon Ragonnet <manonr...@gmail.com> wrote:

Edit: it's a Linux server not a cluster. BEAST2 runs no problem as long as the xml doesn't require any packages.

--
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.

[Manon Ragonnet]

Dear all,

I have located and fixed the error and thought the solution might be of interest to others.

I had installed the packages using addonmanager but there turned out to be two folders in ~/.beast/:

~/.beast/2.3/ and ~/.beast/2.4/. One of them must have been from an earlier install of BEAST2. Deleting both folders and installing the packages again solved the problem.

Thanks,

Manon

[Anwesha Saha]

Hi everyone.

I am not sure if this is the right group to ask for help. I am very new to BEAST Ver2.4.5. I have a small dataset of 59 species (including outgroup) of 3 genes. I want to date the tree with 4 calibration points.Besides,  I want to calculate rate shift and diversification rate with BDSKYLINE plugin in BEAST ver 2.4.5. I am trying to prepare the XML file in BEAUTI2. I am not sure about the following parameters.

1) How to choose the appropriate prior for (become uninfectious rate; Origin; rho; reproductive number)

2) How to decide on the offset values for each parameter mentioned above?

It would be of great help if anyone can share any exercise sheet with an example dataset for such analysis.

Or At least if someone could tell me whom should I write to regarding this or post in an appropriate group?

Waiting eagerly for someone's response.

Cheers

Mimi

[Alexei Drummond]

Hi Mimi,

* offsets should be zero.

* origin can have an almost non-informative prior and a starting value larger than the root of the tree.

* reproductive number is expected to be larger than 1 and unlikely to be larger than 10 for most situations. But you should set a prior based on what you know about your system.

* become-uninfectious-rate is the recovery rate. It should have a value that is the inverse of the average duration of infection (in the same units as your tip dates are expressed in).

* rho should be between 0 and 1 and represents the fraction of the population sampled at the final sampling time.

Alexei