Using starBEAST for molecular dating

[Kunal Arekar]

Hi All

I tried using starBEAST for obtaining a time tree from multilocus nuclear data. I tried more than 20 times but it just will not run. There will be some or the other error. And if I try to correct for error, a new one crops up. 
Below is the error I got after the last analysis I tried to run, after this I gave up, could not run any more analysis

Start likelihood: -Infinity after 11 initialisation attempts

P(posterior) = -Infinity (was NaN)

P(speciescoalescent) = 94.1308418134601 (was NaN)

P(SpeciesTreePopSize.Species) = 63.065611315535484 (was NaN)

P(treePrior.t:ABCA1) = 31.065230497924613 (was NaN)

P(prior) = -Infinity (was NaN)

P(YuleModel.t:Species) = 13.989377215277079 (was NaN)

P(YuleBirthRatePrior.t:Species) = -3.5923451625743414 (was NaN)

P(KappaPrior.s:ABCA1) = -1.8653600339742873 (was NaN)

P(KappaPrior.s:BCHE) = -1.8653600339742873 (was NaN)

P(KappaPrior.s:BDNF) = -1.8653600339742873 (was NaN)

P(KappaPrior.s:DMRT1) = -1.8653600339742873 (was NaN)

P(KappaPrior.s:ERC2) = -1.8653600339742873 (was NaN)

P(KappaPrior.s:FAM) = -1.8653600339742873 (was NaN)

P(KappaPrior.s:FES) = -1.8653600339742873 (was NaN)

P(KappaPrior.s:MAP) = -1.8653600339742873 (was NaN)

P(popMean.prior) = 5.208719930051549 (was NaN)

P(MeanRatePrior.c:ABCA1) = 0.0 (was NaN)

P(Afro-Asian.prior) = -97.0048815145909 (was NaN)

P(Mesopithecus.prior) = -Infinity (was NaN)

P(likelihood) = NaN (was NaN)

P(treeLikelihood.ABCA1) = NaN (was NaN)

P(treeLikelihood.BCHE) = NaN (was NaN)

P(treeLikelihood.BDNF) = NaN (was NaN)

P(treeLikelihood.DMRT1) = NaN (was NaN)

P(treeLikelihood.ERC2) = NaN (was NaN)

P(treeLikelihood.FAM) = NaN (was NaN)

P(treeLikelihood.FES) = NaN (was NaN)

P(treeLikelihood.MAP) = NaN (was NaN)

java.lang.Exception: Could not find a proper state to initialise. Perhaps try another seed.

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)

If anyone knows how can I use starBEAST (Instead of BEAST) for getting a time tree from my data, please let me know.

*PS: I am using the starBEAST template embedded in BEAST v2.1.3.

[Huw A. Ogilvie]

Hi Kunal,

v2.1.3 is an extremely old version of BEAST2 by now. I would recommend trying to set up and run the analysis again in the latest version, and letting us know if it still does not initialize.

- Huw

[Kunal Arekar]

Hi Huw

Thanks for the suggestion. I tried running the analysis in the latest version (v2.4.7). But the same error is repeating.

--
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/3pUwoskDNOk/unsubscribe.
To unsubscribe from this group and all its topics, 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.

--

Kunal Arekar
Dr. Karanth's lab

CES, IISc

Bangalore

+918861532520

[Kunal Arekar]

Here is the error...

Start likelihood: -Infinity after 10 initialisation attempts

P(posterior) = -Infinity (was -Infinity)

P(speciescoalescent) = 136.82945190669415 (was 136.82945190669415)

P(SpeciesTreePopSize.Species) = 81.98415701140536 (was 81.98415701140536)

P(treePrior.t:ABCA1) = 54.8452948952888 (was 54.8452948952888)

P(prior) = -Infinity (was -Infinity)

P(YuleModel.t:Species) = 20.016822533206483 (was 20.016822533206483)

P(YuleBirthRatePrior.t:Species) = -5.058573886139971 (was -5.058573886139971)

P(KappaPrior.s:ABCA1) = -1.8653600339742873 (was -1.8653600339742873)

P(KappaPrior.s:BCHE) = -1.8653600339742873 (was -1.8653600339742873)

P(KappaPrior.s:BDNF) = -1.8653600339742873 (was -1.8653600339742873)

P(KappaPrior.s:DMRT1) = -1.8653600339742873 (was -1.8653600339742873)

P(KappaPrior.s:ERC2) = -1.8653600339742873 (was -1.8653600339742873)

P(KappaPrior.s:FAM) = -1.8653600339742873 (was -1.8653600339742873)

P(KappaPrior.s:FES) = -1.8653600339742873 (was -1.8653600339742873)

P(KappaPrior.s:MAP) = -1.8653600339742873 (was -1.8653600339742873)

P(popMean.prior) = 6.3911290360434165 (was 6.3911290360434165)

P(MeanRatePrior.c:ABCA1) = 0.0 (was 0.0)

P(MeanRatePrior.c:BCHE) = 0.0 (was 0.0)

P(MeanRatePrior.c:BDNF) = 0.0 (was 0.0)

P(MeanRatePrior.c:DMRT1) = 0.0 (was 0.0)

P(MeanRatePrior.c:ERC2) = 0.0 (was 0.0)

P(MeanRatePrior.c:FAM) = 0.0 (was 0.0)

P(MeanRatePrior.c:FES) = 0.0 (was 0.0)

P(MeanRatePrior.c:MAP) = 0.0 (was 0.0)

P(ucldStdevPrior.c:ABCA1) = 0.819364145120802 (was 0.819364145120802)

P(ucldStdevPrior.c:BCHE) = 0.819364145120802 (was 0.819364145120802)

P(ucldStdevPrior.c:BDNF) = 0.819364145120802 (was 0.819364145120802)

P(ucldStdevPrior.c:DMRT1) = 0.819364145120802 (was 0.819364145120802)

P(ucldStdevPrior.c:ERC2) = 0.819364145120802 (was 0.819364145120802)

P(ucldStdevPrior.c:FAM) = 0.819364145120802 (was 0.819364145120802)

P(ucldStdevPrior.c:FES) = 0.819364145120802 (was 0.819364145120802)

P(ucldStdevPrior.c:MAP) = 0.819364145120802 (was 0.819364145120802)

P(Afro-Asian.prior) = -100.6089383058432 (was -100.6089383058432)

P(Mesopithecus.prior) = -Infinity (was -Infinity)

P(likelihood) = NaN (was NaN)  **

P(treeLikelihood.ABCA1) = NaN (was NaN)  **

P(treeLikelihood.BCHE) = NaN (was NaN)  **

P(treeLikelihood.BDNF) = NaN (was NaN)  **

P(treeLikelihood.DMRT1) = NaN (was NaN)  **

P(treeLikelihood.ERC2) = NaN (was NaN)  **

P(treeLikelihood.FAM) = NaN (wjava.lang.RuntimeException: Could not find a proper state to initialise. Perhaps try another seed.

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)

Fatal exception: Could not find a proper state to initialise. Perhaps try another seed.

as NaN)  **

P(treeLikelihood.FES) = NaN (was NaN)  **

P(treeLikelihood.MAP) = NaN (was NaN)  **

Fatal exception: Could not find a proper state to initialise. Perhaps try another seed.

BEAST has terminated with an error. Please select QUIT from the menu.

(Sorry for the two separate mails. I accidentally pressed enter button).

Is there some problem with my calibration. I am not able to figure out the problem here.

Thanks

Kunal

[Kunal Arekar]

Hi All

In relation to the same issue described above when I run the starBEAST analysis without the calibration prior, it shows no error. 

But as soon as I introduce a calibration prior, it gives an initialization error. 

I need to resolve this as soon as possible. All suggestions are appreciated.

Thanks in advance
Kunal

[Carlo Pacioni]

Hi Kunal,

It seems that BEAST can't calculate the likelihood with the calibration you are providing. If you have a starting value for the age of the clade you provided the calibration for, it may be possible that this is not compatible with your prior. 

carlo

 

--
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.

[Kunal Arekar]

Hello Everyone

Thank you all those who gave suggestions. Unfortunately, the problem is not yet resolved.

I would like to point out here that when I used log-normal distribution and exponential distribution with my fossil calibration prior, I get this error.
Do you think this is because of some bug in BEAST 2.4.7? 
I request the developers of BEAST 2 to look into this matter and help me out with the analysis.

Cheers!!
Kunal

To unsubscribe from this group and all its topics, 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.

--

Kunal Arekar
Dr. Karanth's lab

CES, IISc

Bangalore

+918861532520

--

Kunal Arekar
Dr. Karanth's lab

CES, IISc

Bangalore

+918861532520

--

Kunal Arekar
Dr. Karanth's lab

CES, IISc

Bangalore

+918861532520

--
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.

[Huw A. Ogilvie]

Hi Kunal,

Could you attach your XML file, or email it to me, so I can try and work out what the problem with calibrations is?

- Huw

[Huw A. Ogilvie]

Thanks for the XML file Kunal.

The MRCA prior should be applied to the species tree, not one of the gene trees. In XML file you sent me, it was applied to the "ABCA1" gene tree, which is why the initialization failed. I changed the prior to apply to the species tree, and now the chain starts okay.

- Huw

[Kunal Arekar]

Hi Huw

The problem is solved. I put the prior on the species tree, now it's working.

But I did apply the prior on the species tree before, but it did not work then.

Anyways now it's working, so everything is good.

Thanks for the suggestion

Cheers!