I’m trying to make a pipeline in python that will make use of MGL Tools (https://anaconda.org/bioconda/mgltools) as part of the process.

I have a bunch of mutated protein structures that I’d like to dock ligands against, using autodock vina.

I’m working with about 16 000 mutant structures, so am trying to process them using MPI, in parallel.

For each structure, I need to prepare it using one of the scripts from MGLTools, to add partial charges to the protein (Gasteiger charges).

I’ve installed the mgltools package to my conda environment.

However, when I try to run a job on the cluster (PBS Pro), I’m not able to access my conda environment that I created.
Therefore, I can’t run the python scripts from MGL Tools.

Any thoughts/ advice?

Regards

Jeremy Burgess

Here follows the code I’m using:

module purge
module load chpc/BIOMODULES openmpi
module load chpc/python/anaconda/3-2021.11

echo $(module list)

conda init bash
conda list

eval "$(conda shell.bash hook)"
/apps/chpc/chem/anaconda3-2021.11/bin/conda activate open_docking_env

cd /mnt/lustre/users/jburgess1/mutation_tests/chatgpt/April/

NP=$(cat ${PBS_NODEFILE} | wc -l )
mpirun --mca btl '^openib' -np $NP --machinefile $PBS_NODEFILE python 4_Mpi_Initialize_MGL_TOOLS.py

conda deactivate