Error - Coupled MCMC with Speedemon?

[Monica Ó Fathaigh]

Hi Beast 2 users, 

Does CoupledMCMC work with Speedemon/snapper? I ran a speedemon analysis with 40 samples and ~5000 snps. Using 80 threads it took 4 weeks to cover about 880000 steps. 

To speed things up, I reduced the SNPs to ~3000 and tried 2 chains with CoupledMCMC. 

I got this error message:

 <distribution id='prior' spec='CompoundDistribution'>
                      <distribution id='SPEEDEMONYuleSkylineCollapse.t:EndMuelHaye_SPLITSTREE_5Taxa_8sub_NSW' spec='speedemon.YuleSkylineCollapse'>
                          <tree id='Tree.t:EndMuelHaye_SPLITSTREE_5Taxa_8sub_NSW' spec='beast.base.evolution.tree.Tree'>
                              <taxonset id='TaxonSet.EndMuelHaye_SPLITSTREE_5Taxa_8sub_NSW' spec='TaxonSet'>
                                  <alignment id='snapper.EndMuelHaye_SPLITSTREE_5Taxa_8sub_NSW' spec='snapper.Data'>

        at coupledMCMC.CoupledMCMC.initRun(CoupledMCMC.java:248)
        at coupledMCMC.CoupledMCMC.run(CoupledMCMC.java:309)
        at beastfx.app.beast.BeastMCMC.run(Unknown Source)
        at beastfx.app.beast.BeastMain.main(Unknown Source)
        at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
        at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(Unknown Source)
        at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(Unknown Source)
        at java.base/java.lang.reflect.Method.invoke(Unknown Source)
        at beast.pkgmgmt.launcher.BeastLauncher.run(Unknown Source)
        at beast.pkgmgmt.launcher.BeastLauncher.main(Unknown Source)
Caused by: beast.base.parser.XMLParserException:
Error 110 parsing the xml input file

validate and intialize error: java.lang.RuntimeException: Duplicate taxon found in alignment: EmuelNNSWSEQLD_Wollumbin_NSW1152536

Error detected about here:
  <beast>
      <run id='HeatedChain' spec='coupledMCMC.HeatedChain'>
          <coupledLogger id='CoupledLogger' spec='coupledMCMC.CoupledLogger'>
              <model id='posterior' spec='CompoundDistribution'>
                  <distribution id='prior' spec='CompoundDistribution'>
                      <distribution id='SPEEDEMONYuleSkylineCollapse.t:EndMuelHaye_SPLITSTREE_5Taxa_8sub_NSW' spec='speedemon.YuleSkylineCollapse'>
                          <tree id='Tree.t:EndMuelHaye_SPLITSTREE_5Taxa_8sub_NSW' spec='beast.base.evolution.tree.Tree'>
                              <taxonset id='TaxonSet.EndMuelHaye_SPLITSTREE_5Taxa_8sub_NSW' spec='TaxonSet'>
                                  <alignment id='snapper.EndMuelHaye_SPLITSTREE_5Taxa_8sub_NSW' spec='snapper.Data'>

        at beast.base.parser.XMLParser.initBEASTObjects(Unknown Source)
        at beast.base.parser.XMLParser.parseFragment(Unknown Source)
        at coupledMCMC.CoupledMCMC.initRun(CoupledMCMC.java:221)

There are no taxon duplicates in the xml file (see here: EndMuelHaye_SD6_5Taxa_8sub_3146snps_ind.xml) and the same dataset works as a normal speedemon analysis. 

Cheers,

Monica

[B Godwin]

Hi Monica,

Did you ever find a resolution to this error? I just encountered it myself. It runs fine with normal MCMCs in an otherwise-identical .XML file.

[higg...@gmail.com]

Hi Brady, 

I tried to replicate the problem by adding an MRCA prior to a snapper example file (https://github.com/rbouckaert/speedemon/blob/master/examples/leache.xml) and it starts without problem both with MCMC and with Coupled MCMC.

The XML file Monica referred to expired. Maybe you or Monica can send me the XML so I can investigate what is causing the problem.

Cheers, Remco

[Monica Ó Fathaigh]

Hi Remco

Here is the XML: https://drive.google.com/file/d/1OYH15IfyMZ5pn6a6tztOQvn7OeVoTrTC/view?usp=sharing

I had another attempt. I can get the XML to run from the Windows desktop app (only when N=33, it gives an error when N=65). When I ran on the command line on a linux server I got the same error as above (where it claims there is a duplicate taxon).

The only way I can convert my XML is if I manually edit in a text editor as per the recommendation here (https://www.beast2.org/2020/01/14/metropolis-coupled-mcmcmc3-works.html). The converter app doesn't launch when I try to open it in beauti. Also, I had a look at your example XML and can't see the specifications for coupledMCMC.

Thanks for your help.

Cheers

Monica

[higg...@gmail.com]

Hi Monica,

Thanks for the XML -- I tried to launch it on OS X and Linux and in both cases it starts without problems, so I suspect it has something to do with the way BEAST is installed.

Do you have the list of packages available on the Linux server: BEAST should print it out at the start of the run just after it shows which file it runs, something like

Loading package BICEPS v1.1.2 feast v10.3.0 NS v1.2.2 bModelTest v1.3.3....

Cheers, Remco

[Monica]

Hi Remco,

I'm using Beast v2.7 with these packages installed: BEASTlabs.v2.0.2, snapper, ORC, CoupledMCMC.v1.2.1, starbeast3, MODEL_SELECTION, BICEPS v1.1.2, speedemon, SNAPP, SA.v2.1.1, DensiTree.

I tested on two servers. One has beast installed manually on a common directory. On the other I installed beast on my own directory using a dedicated environment in miniconda. Both had the same error...

--
You received this message because you are subscribed to a topic in the Google Groups "beast-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/beast-users/5krcsvkbey4/unsubscribe.
To unsubscribe from this group and all its topics, send an email to beast-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/beast-users/21f44576-93e0-4da3-83d3-fdaacb02b4d2n%40googlegroups.com.

[Remco Bouckaert]

Thanks Monica,

DensiTree should not be a seperate BEAST 2 package, it is just part of the distribution, so I assume this was a mistake? I am still not able to replicate the problem you encountered. The only thing I can think of is that something did not go quite right when installing packages. If so, removing all packages, then re-installing the ones you need might fix the problem.

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 view this discussion on the web visit https://groups.google.com/d/msgid/beast-users/CACMjf59%3DyxvLc%3DQPweEZ-HhTeDGCDG5Kz8CF1MRVy8rNxhzpfg%40mail.gmail.com.