Computational Drug Discovery and Design

vina --config vina.conf --cpu 1 --ligand \$LIGAND \

--out ${outdir}/\${lnam}_out.pdbqt \

--log ${outdir}/\${lnam}_log.dat

END

# launch the job

qsub -R y -V -t 1-$nlig -tc $nconc $jid

In the above, SGE_TASK_ID is a unique identifier attributed

to each processor and initialized by the system at the beginning of

the program’s execution. The number of simultaneously running

jobs is limited to nconc (option –tc for the qsub command) in order

to keep the cluster available for other users. An example of the input

file vina.conf is given below:

# specify cpu, ligand, out & log on the command line

receptor = receptor.pdbqt

center_x = 4.531

center_y = -23.703

center_z = 66.949

size_x = 42.0

size_y = 42.0

size_z = 45.0

num_modes = 20

exhaustiveness = 100

energy_range = 5

Adapt the names of files and the grid size/center coordinates to

your case.

3. At the end of the docking process, rank your molecules for each
   structure according to their Vina scores (in kcal/mol) in the
   output log files and a selected threshold value (seeNote 8). You
   can also create a script to extract and rank only the first
   conformer for each molecule (presumed to be the most stable
   conformer), as in the example below:

#!/bin/sh

outdir = "results" (adapt this line to your case)

nlig=‘ls -1 $outdir/*log.dat | wc -l‘

i=1

rm -f x_tmp

ls -1 $outdir/*log.dat |

while read fnam; do

nrj=‘grep ’^ 1’ $fnam | cut -c10-20‘;

164 Gre ́gory Menchon et al.