GNU-Parallel-Vina: GNU Parallel Based Implementation of Autodock Vina

---

GNU-Parallel-Vina is an open-source tool that parallelizes AutoDock Vina, aimed at drastically reducing virtual screening times by utilizing compute clusters or networks of computers. Developed using GNU Parallel, it supports both shared and distributed memory environments. The primary goal is to accelerate the screening of ligand datasets, distributing protein-ligand docking tasks across multiple CPU cores and NODES for simultaneous processing.

Requirements

GNU Parallel (can be installed via package managers like apt, yum, or brew).

Usage

1. Prepare Files:

   • Convert your target receptor (i.e., protein) and all ligands into PDBQT format using traditional tools.
   • Place all ligand files in the Ligand directory.
   • Place your configuration file (conf.txt) in the Vina directory.
   • Place your target receptor in the src directory.

2. Adjust Job Settings:

   • Edit runParallelVINA.bash and adjust JOBS_PER_NODE according to your needs.
   • When using a PBS scheduler, edit the runParallelVINA.bash and add the --slf $PBS_NODEFILE flag to access all allocated nodes.

3. Running the Program:

   • Run the script by either typing ./runParallelVINA.bash in the command line or submitting the submit_vina_job.pbs file via qsub.

4. Output Files:

   • All output files will be placed in the Output directory:
     • job.log contains detailed logs of the parallel job execution.
     • ParallelVINA.log contains a summary of the parallel processing.
     • SortedResult contains binding affinities of all ligands in sorted order.
     • <ligand_name>.pdbqt.pdbqt and <ligand_name>.pdbqt.txt files contain detailed results of individual ligands.

5. Processed Ligands:

   • The ProcessedLigand directory contains:
     • <ligand_name>.pdbqt files of individual ligands.

6. Failed Jobs:

   • Edit runParallelVINA.bashand use the --resume-failed flag to re-run failed jobs.

Acknowledgement

This project is a fork of the original MPI-Vina.