run_get_phylomarkers_pipeline.sh stops without feedback

[TT1972]

Hi, I am using get_phylomarkers on a HPC cluster using slurm. On the cluster I can't use docker and have to use apptainer instead.
The sbatch file looks like this:

!/bin/bash

#SBATCH -p defq
#BSUB -J get_phylom
#BSUB -o stdout.%J
#BSUB -e stderr.%J

module load apptainer
apptainer run --bind /home/thuenen/work:/home/you/get_homPhy/ apptainer/get_phylomarkers_latest.sif "/home/you/get_homPhy/get_phylo.sh"

The get_phlyo.sh script looks like this:

#!/bin/bash

cd /home/you/get_homPhy/Mesorhizobium/GBFF_homologues/core_B_Mesorhizobia
run_get_phylomarkers_pipeline.sh -R 1 -t DNA

The problem is, that the run_get_phylomarkers_pipeline.sh script is not running through completely, but simply stops at some point, without giving any error message. I have attached the log file.

Is there a way to find out, what is happening, and why the script stops?

Best regards,
Torsten

get_phylomarkers_run_AIR1tDNA_k1.5_m0.7_Thigh_18Mar25.log

[TT1972]

I just found the wastral_usertree_concat_model_selection.log which shows some errors:

Creating fast initial parsimony tree by random order stepwise addition...
0.92 seconds, parsimony score: 528392 (based on 71346 sites)
Alignment was printed to wastral_usertree_concat_model_selection.varsites.phy
ERROR: For your convenience alignment with variable sites printed to wastral_usertree_concat_model_selection.varsites.phy
ERROR: Invalid use of +ASC because of 1 invariant sites in the alignment

The full log is attached.

Best regards,
Torsten

wastral_usertree_concat_model_selection.LOG

[vinuesa]

Hi Torsten @TT1972, thanks for reaching out. From the wastral logfile it seems that the problem lies in the concatenated alignment, which should contain only variable sites (invariant columns are removed), but a single invariant column remains for some unknown reason. This causes iqtree to complain, as the use of +ASC model can only be fitted to alignments containing only variable sites. This may be caused perhaps by unresolved residues in the sequence of a particular locus. Could you inspect the concatenated alignment to search for the insulting invariant site, or share it with me? It could be also helpful to repeat the run enabling verbose debugging with -D 1 or -D 2.

[TT1972]

Hi Pablo @vinuesa,
please find attached the alignment. I couldn't find any invariant column, though.
I will repeat the run with verbose debugging and let you know the results.

wastral_usertree_concat_model_selection.varsites.phy.zip

[TT1972]

I just realized, that the phy file wasn't to one iqtree used. Here is the actual alignment. It is one bp longer than the phy file. I also have attached the tree file used in the iqtree command.

I ran iqtree with the same options used in the script. The job on the HPC cluster failed with ExitCode 2 (Same as the get_phylomaker script is giving me). I have alltached the log as well.

wastral_top150geneTrees.zip

iqtree_log.zip

concat_cdnAlns.zip

[TT1972]

Found the position which is causing the problem. It's position 62888. Here sequence 30 has a degenerate nucleotide code M in its sequence.

In case you still need the log file with verbose debugging, I have attached it here

get_phylomarkers_run_AIR1tDNA_k1.5_m0.7_Thigh_20Mar25.log