crystallography, and molecular modeling approaches [1–16], and
more recently through the cryo-electron microscopy [17]. There is
no unique protocol or solution for dealing with a drug discovery
problem but the general workflow of a virtual ligand screening
campaign can be compared to a “funnel strategy” and usually starts
with a high-throughput docking experiment, with filtered and
enriched ligand libraries and a set of representative 3D protein
structures determined with the previously mentioned techniques.
See Fig.1 for a general schematic overview of this procedure.
Computational docking allows finding the best position, orienta-
tion and theoretical affinity of a ligand in the binding pocket of a
target receptor [18–21]. However, docking methods also face
major issues, especially considering that the mechanisms of pro-
tein–ligand recognition and binding are dynamic processes.
Although flexible docking protocols easily predict different ligand
conformations, the docking protocols and output results cannot
account for full protein flexibility, induced-fit, and solvent effects,
nor for protein–ligand complex stability and accurate binding free
energy estimation. In that situation, flexible docking technique canFig. 1A general schematic illustration of the funnel-like approach to drug discovery. This chapter deals with
virtual screening, representing the first step of this method
146 Gre ́gory Menchon et al.