BEAST2 run terminates (partII)

[Susana Freitas]

Dear all,
I posted previously with a error on invalid jarfile, but I already corrected it (to run beast2 on a linux terminal, we have to run the bash file with the input file: $ ./beast MyBeautiFile.xml).

I am trying to run an alignment to calculate the tMRCA. I made the beauti input file, but every time I run it the run terminates abruptly without even starting.
This is the error I get

                               Thanks to:
          Roald Forsberg, Beth Shapiro and Korbinian Strimmer

Random number seed: 1439828489769

File: eagora.xml seed: 1439828489769 threads: 1
Alignment(all_cytb_BEAST_clean3)
  267 taxa
  919 sites
  399 patterns

Alignment(all_ND4)
  267 taxa
  839 sites
  378 patterns

  UCRelaxedClockModel: using 532 rate categories to approximate rate distribution across branches.
Failed to load BEAGLE library: no hmsbeagle-jni in java.library.path
TreeLikelihood(treeLikelihood.all_cytb_BEAST_clean3) uses BeerLikelihoodCore4
  Alignment(all_cytb_BEAST_clean3): [taxa, patterns, sites] = [267, 399, 919]
  UCRelaxedClockModel: using 532 rate categories to approximate rate distribution across branches.
Failed to load BEAGLE library: no hmsbeagle-jni in java.library.path
TreeLikelihood(treeLikelihood.all_ND4) uses BeerLikelihoodCore4
  Alignment(all_ND4): [taxa, patterns, sites] = [267, 378, 839]
WARNING: no up item specified in UpDownOperator
WARNING: the dimension of the parameter is 1 at the start of the run.
         The operator FixMeanMutationRatesOperator has no effect (if this does not change).
===============================================================================
Citations for this model:

Bouckaert RR, Heled J, Kuehnert D, Vaughan TG, Wu C-H, Xie D, Suchard MA,
  Rambaut A, Drummond AJ (2014) BEAST 2: A software platform for Bayesian
  evolutionary analysis. PLoS Computational Biology 10(4): e1003537

Heled J, Drummond AJ (2012) Calibrated Tree Priors for Relaxed Phylogenetics
  and Divergence Time Estimation. Systematic Biology 61(1):138-149.

Drummond AJ, Ho SYW, Phillips MJ, Rambaut A (2006) Relaxed Phylogenetics and
  Dating with Confidence. PLoS Biol 4(5): e88

===============================================================================
java.lang.NullPointerException
    at beast.evolution.speciation.CalibratedYuleModel.getCorrection(Unknown Source)
    at beast.evolution.speciation.CalibratedYuleModel.calculateTreeLogLikelihood(Unknown Source)
    at beast.evolution.speciation.SpeciesTreeDistribution.calculateLogP(Unknown Source)
    at beast.core.util.CompoundDistribution.calculateLogP(Unknown Source)
    at beast.core.util.CompoundDistribution.calculateLogP(Unknown Source)
    at beast.core.State.robustlyCalcPosterior(Unknown Source)
    at beast.core.MCMC.run(Unknown Source)
    at beast.app.BeastMCMC.run(Unknown Source)
    at beast.app.beastapp.BeastMain.<init>(Unknown Source)
    at beast.app.beastapp.BeastMain.main(Unknown Source)
    at beast.app.beastapp.BeastLauncher.main(Unknown Source)

I tried to run the example xml files and they seemed to work, so the problem is on my xml.
Do any of you have any idea of what might be happening?

Thanks in advance,
Best!
Susana

[Remco Bouckaert]

Hi Susana,

Is it possible that you did not specify a distribution for the clade that you are interested in and that the clade is not monophyletic?

The Calibrated Yule prior picks up these distributions, but if none is specified this (admittedly not very informative) error appears when the clade is not monophyletic.

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 http://groups.google.com/group/beast-users.
For more options, visit https://groups.google.com/d/optout.

[Susana Freitas]

Hello,
Thanks for the prompt answer!

I have set the ucedMean for one of the markers (with an uncorrelated exponential relaxed clock mean) as:
mean: 0.0226
sigma: 0.0031
and offset: 0.0
so that it could vary between the mutation interval previously estimated for this marker. Is that what you mean?

Clades are monophyletic, but even if they weren't I am not sure if it ran long enough to make any tree and understand which are the monophyletic clades or not (assuming it was not referring to repeated sample names in different traits, which there aren't).
I kept trying different priors, and if I change to Yule model (not the calibrated version), it works. However, I thought since I was using only one calibration (I am estimating only one marker subs rate) I assumed the calibrated Yule would perform better than the non-calibrated.

Do you think this change would be significant?

Many thanks!
Susana

[Remco Bouckaert]

HI Susana,

The kind of calibrations that the Calibrated Yule prior deals with are of the form where the monophyly and age of a clade are specified by a prior (you can add such calibration in BEAUti by clicking the ‘+’ button at the bottom of the priors panel).

If you did not specify such a calibration, but only have timing information in the form of a distribution on the rate (as you mentioned), the calibrated Yule is not appropriate, and you should use the standard Yule instead.

Cheers,

Remco

[Susana Freitas]

My mistake. The runs finished smoothly, many thanks for the help (and software).

Cheers,
Susana