echo $nrj $fnam>>x_tmp;
i=‘expr $i + 1‘;
done
sort -n –k1 x_tmp>sorted_NRG
rm x_tmp
The output file sorted_NRG contains the sorted list of the
ligands, best on top. Inspect the intermolecular contacts and select
a number of candidates for the following steps. Then, run MD
simulation to validate the selected protein–ligand complexes.
Visualization of the OpenEye docking results can be done with
Vida, e.g., with the command "vida docked.oeb.gz". One can also
save the structures in the mol2 format (as one single file) to view
them with Pymol (as individual models).
3.6 Validation
of Complexes by MD
Simulations
- Putative complex structures resulting from docking studies
should be validated from the point of view of their stability.
The MD simulation performed even for a relatively short
stretch of time (10–20 ns) can give a wealth of information
about the behavior of the ligand within the binding site, as well
as about the nature of the intermolecular contacts.
The preparation of input data for MD simulation of com-
plex structures is similar to the procedure described above for
the receptor. The difference is that a new molecule (ligand) has
to be added to the system and we need its topology. In general,
it will require the use of a different force field. In what follows
we assume that ligands are small organic molecules which can
be treated by the general Amber force field (GAFF) [54].
We start with a PDB file of a ligand molecule. Make sure
that the ligand is protonated. Use the antechamber utility to
generate the first of the tleap input files:
antechamber –nc 0 –rn LIG –I lig.mol2 –fi mol2 –o lig.prep \
-fo prepc –c bcc –s 2
The meaning of the options is as follows: ns¼net charge, rn¼
residue name, i¼input file, fi¼format of the input file, o¼output
file, fo¼format of the output file (prepc or prepi), c¼charge
model, s¼output verbosity. When the lig.prep file is ready, we
check if all force field parameters are available:
parmchk -i lig.prep -f prepc -o lig.frcmod
The contents of the output frcmod file should be examined,
and if problems occur, they should be fixed before continuing (see
Note 9).
Molecular Dynamics in Virtual Screening 165