Computational Drug Discovery and Design

(backadmin) #1
to launch an MD simulation, to analyze the MD trajectories, to
extract representative receptor structures and use them in docking
of a library of filtered molecules and finally to validate the putative
complexes by calculating their stability, detecting interatomic inter-
actions and by computing ligand affinity parameters.
In this chapter we describe the general virtual ligand screening
workflow by implementing docking coupled with MD simulations.
Each step, from the preparation of the receptor and ligand input
files to the validation of the complexes is described in detail and
gives the reader an overview of a complete and general strategy
which is broadly used by computational biologists to find new
binding molecules with pharmacological interest. We especially
emphasize the need to use the information on molecular flexibility
during a virtual screening procedure, which is mandatory to fully
explain and predict a binding event between a receptor and a
putative binding molecule.

2 Materials


The existing MD and VS software runs on a variety of platforms,
but in this chapter we will focus on the programs we have used in
our laboratory. Some of the other existing options will be given in
references, but this is not meant to be an exhaustive review. The
calculations have been performed on Desktop Workstations
operating under the Linux OS (Centos 6.6), equipped with NVidia
Titan Black GPU cards (http://www.nvidia.com/gtx-700-

Fig. 2Example of stability analysis (RMSD vs time) for a stable molecule (blue)
and an unstable one (red) within a target protein, along a 10 ns MD trajectory

148 Gre ́gory Menchon et al.

Free download pdf