generax terminate after throwing an instance of 'LibpllException'

[Emmanuelle Morin]

Hello,

I have installed GeneRax from Bioconda on a CentOS 8 cluster.

I am trying to use GeneRax to infer a species tree from 238 genes.

Here is my command :

mpiexec -np 40 generax --families families_boletales.txt --species-tree MiniNJ --si-strategy HYBRID --rec-model UndatedDTL --per-family-rates --prune-species-tree --si-estimate-bl --si-quartet-support --prefix 238_protz

It fails after or during Species BL estimation :

[00:12:13] Reconciling gene trees with the species tree...

[00:12:20] [Species BL estimation] Infering reconciliation scenarios

[00:12:20] [Species BL estimation] Infering branch lengths from gene trees

[00:12:20] [Species BL estimation] Done

terminate called after throwing an instance of 'LibpllException'

  what():  Could not load open newick file __random__terminate called after throwing an instance of 'LibpllException'

  what():  Could not load open newick file __random__

[node008:2954214] *** Process received signal ***

[node008:2954214] Signal: Aborted (6)

[node008:2954214] Signal code:  (-6)

Any idea what is going wrong ?

Thanks,

Emmanuelle

[Benoit Morel]

Dear Emmanuelle,

I apologize for answering so late, the google group notifications of the last two weeks ended up in my spam folder...

This reminds me a bug that should have been fixed. Can you confirm that you only gave as input the gene alignments, and that the gene trees were inferred by GeneRax? Could you please send me the file families_boletales.txt?

I will investigate on my machine, and if I can't reproduce the problem, I'll ask you if you can share the whole dataset (if possible).

Note that the rooted species tree should already have been computed, you can find it under species_trees/inferred_species_tree.newick (to be certain, please check that it's different than the file species_trees/starting_species_tree.newick).

Best,

Benoit

[Emmanuelle Morin]

Hello,

Yes I only gave the gene alignments as input. Please find attached the families_boletales.txt file.

The rooted species tree is indeed computed but I was worried that the tree was not the final one.

Thanks,

Emmanuelle

[Benoit Morel]

Hello Emmanuelle,

I think I managed to fix your issue. I will let you know when the corresponding bioconda package is available (I started the process, but it sometimes takes a few days).

The crash occurred while trying to compute the quartet-based support values. As a result, the species tree inference successfully finished, but the support values should be missing. As I already wrote, the crash does not affect the inferred tree topology.

Many thanks for the report!

Best,

Benoit