This is a test version of Biostars. For the public version, visit https://www.biostars.org.
How to run beast.jar file using Slurm?

I have installed BEAST using conda in HPC.

$conda install -c bioconda beast

Now I'm trying to run it through Slurm script using 10 million generation

#!/bin/sh
#SBATCH --job-name=Beast_10Mn             # Job name
#SBATCH --ntasks=16                       # Run on the no. of CPU
#SBATCH --ntasks-per-node=4               # cores to spread across distinct nodes
#SBATCH --output=BST_log_%j               # Standard output log
#SBATCH --error=job.%J.err                # Standard error log


for i in dipTime_10mn.xml
do
echo "${i}"
java -jar /lib/beast.jar -threads 16 "${i}"
done;

But it shows Error: Unable to access jar file /lib/beast.jar

It will be an immense help if someone kindly helps.

Thanks,

beast phylogeny jar hpc slurm
• 3,997 views • link
written by kabir.deb • kabir.deb

If you are using SLURM on a cluster it is unlikely that you have permissions to write to /lib directory. Your beast.jar must be installed in a local directory by conda. Find and replace that path.

You would also want to use the for loop outside to set individual SLURM jobs up to make this efficient. Submitting a single job with a for loop inside it is not taking advantage of your cluster resources.

• 0 views • link
written by GenoMax GenoMax

2 answers

SOLVED

#!/bin/sh
#SBATCH --job-name=R4_beast            # Job name
#SBATCH --ntasks=2                     # Run on the no. of CPU
#SBATCH --ntasks-per-node=2            # cores to spread across distinct nodes
#SBATCH --output=R4_%j.log             # Standard output log
#SBATCH --error=R4_%J.err              # Standard error log

for i in *.xml
do
echo "${i}"
/home/d.kabiraj/software/BEAST2.6/beast2.6/bin/XML_input/R4/beast/bin/beast "${i}"
done;

Instead of jar file I used sh file with multiple XML input.

However, still I have a problem I can not connect with the BEAGLE library

• 0 views • link
written by kabir.deb • kabir.deb

If you installed BEAST with conda, it won't be located at /lib/. Besides, BEAST has several wrapper scripts, you don't need to call java -jar.

The conda installer will put the wrapper scripts (applauncher, beast, beauti, loganalyser, logcombiner, treeannotator, treestatin) in the base conda PATH, or in the BEAST env PATH, in case you installed BEAST onto its own environment (which is advisable).

• 0 views • link
written by h.mon h.mon

Log in to answer this question.

More posts like this