Problems running raxml-ng, Aborted, 'core dumped'

[Lucía Moreyra]

Hi!

I usually run the same lines to run raxml-ng (local and on a cluster) without issues:

raxml-ng --all --msa FcC_supermatrix.phy.raxml.rba --bs-metric fbp,tbe --outgroup Echinops-karatavicus_S142 --force perf_threads

or

raxml-ng --msa FcC_supermatrix.phy --tree-constraint constraintree.txt --force perf_threads --model FcC_supermatrix_partition.txt --prefix treesearch100 --seed 91 --tree pars{50},rand{50}

Last time I obtained the following error in the cluster:

terminate called after throwing an instance of 'std::out_of_range'

Aborted (core dumped)

I re-run everything locally and I obtained a similar error:

terminate called recursively

terminate called recursively
terminate called after throwing an instance of 'std::runtime_error'
terminate called recursively
Aborted (core dumped)

I tried to run only a tree search and the same happened, so I assumed the problem was the MSA. I run gene trees with raxml-ng and I didn't have any problems, but when I concatenate all those genes, it doesn't work. I tried upgrading raxml-ng to the latest version (v1.2) and also tried --blopt nr_safe because I saw that it worked for other people having the same issue. In this case, it didn't work either. 

I tried running it with a constrained tree too, and it didn't even find one tree before aborting, so I really don't understand what might be causing it. I'm sure I'm not running out of memory.

I did a tree search in the cluster that did work, so I think the problem is when it starts doing the bootstrap replicates, and if I add a constraint tree, aborts before doing anything.

The link to the msa: https://drive.google.com/file/d/1al_K9EzKmkOslydd9wdry4epxW04vhDy/view?usp=drive_link

I also attached the constraint tree and the outputs from the cluster and locally, to see if you can help me. Thank you in advance!

[Oleksiy Kozlov]

Hi Lucia,

there are actually two separate problems:

1. The unconstrained search on the cluster failed due to buffer overflow bug which has been fixed in
raxml-ng v1.2.0. So please upgrade and it should solve the issue.

2. The problem with constrained search is more tricky and I'm still investigating it. They only
quick workaround I can suggest so far is to drop the constraint (which has only a few nodes, so
probably makes no or minimal difference). Another possibility would be to use the old constraint
checking algorithm ("--extra constraint-old"), but then you cannot use parsimony starting trees, and
it is generally less reliable.

Best,
Oleksiy

On 05.10.23 09:44, Lucía Moreyra wrote:
> Hi!
>
> I usually run the same lines to run raxml-ng (local and on a cluster) without issues:
> raxml-ng --all --msa FcC_supermatrix.phy.raxml.rba --bs-metric fbp,tbe --outgroup
> Echinops-karatavicus_S142 --force perf_threads
> or
> raxml-ng --msa FcC_supermatrix.phy --tree-constraint constraintree.txt --force perf_threads --model
> FcC_supermatrix_partition.txt --prefix treesearch100 --seed 91 --tree pars{50},rand{50}
>
> Last time I obtained the following error in the cluster:

> *terminate called after throwing an instance of 'std::out_of_range'
> *
> *Aborted (core dumped)*

>
> I re-run everything locally and I obtained a similar error:

> *terminate called recursively
> *
> *terminate called recursively

> terminate called after throwing an instance of 'std::runtime_error'
> terminate called recursively

> Aborted (core dumped)*

>
> I tried to run only a tree search and the same happened, so I assumed the problem was the MSA. I run
> gene trees with raxml-ng and I didn't have any problems, but when I concatenate all those genes,
> it doesn't work. I tried upgrading raxml-ng to the latest version (v1.2) and also tried --blopt
> nr_safe because I saw that it worked for other people having the same issue. In this case, it didn't
> work either.
> I tried running it with a constrained tree too, and it didn't even find one tree before aborting, so
> I really don't understand what might be causing it. I'm sure I'm not running out of memory.
>
> I did a tree search in the cluster that did work, so I think the problem is when it starts doing the
> bootstrap replicates, and if I add a constraint tree, aborts before doing anything.
>
> The link to the
> msa: https://drive.google.com/file/d/1al_K9EzKmkOslydd9wdry4epxW04vhDy/view?usp=drive_link
> I also attached the constraint tree and the outputs from the cluster and locally, to see if you can
> help me. Thank you in advance!
>

> --
> You received this message because you are subscribed to the Google Groups "raxml" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to
> raxml+un...@googlegroups.com <mailto:raxml+un...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/raxml/a6fce6f0-d916-4e38-921d-cc79faabfaa3n%40googlegroups.com
> <https://groups.google.com/d/msgid/raxml/a6fce6f0-d916-4e38-921d-cc79faabfaa3n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Lucía Moreyra]

Hi Oleksiy,

Thank you for your response. I will ask the cluster service to upgrade raxml-ng to the latest version. Regarding the problems with the constraint tree, I think I will just skip it. I was only trying a few things and it's not really necessary.

Thank you again!

Lucía.

[Oleksiy Kozlov]

Hi Lucia,

Perfect! I now finally found this nasty bug with constraint trees, the fix will be included in the
next release (v1.2.1).

Best,
Oleksiy

> <https://groups.google.com/d/msgid/raxml/a6fce6f0-d916-4e38-921d-cc79faabfaa3n%40googlegroups.com?utm_medium=email&utm_source=footer <https://groups.google.com/d/msgid/raxml/a6fce6f0-d916-4e38-921d-cc79faabfaa3n%40googlegroups.com?utm_medium=email&utm_source=footer>>.

>
> --
> You received this message because you are subscribed to the Google Groups "raxml" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to
> raxml+un...@googlegroups.com <mailto:raxml+un...@googlegroups.com>.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/raxml/d68dc20f-c66c-450c-9477-66255adbb49en%40googlegroups.com
> <https://groups.google.com/d/msgid/raxml/d68dc20f-c66c-450c-9477-66255adbb49en%40googlegroups.com?utm_medium=email&utm_source=footer>.