Chemistry Reference
In-Depth Information
Combination of flexible pocket models, shape and pharmacophore matching
(hydrogen bridges, solvent molecules among others) may contribute to progress
for receptor-based drug design
via
ligand adaptation. This should help avoid, to a
certain degree, extensive usage of ligand poses [42-90].
Molecular docking and high-throughput virtual screening should certainly
continue playing increasingly important roles in the coming decades.
HOMOLOGY MODELING
High-quality 3-D structures of the protein target are important for ligand and high-
throughput docking (HTD) in the drug design and discovery process. There is a
continuous increase in the already large number of proteins available in the
Protein Data Band (PDB). Despite this increase, numerous important drug targets
do not have 3-d structures. As an example of this exception, we can include
membrane proteins with their crystallization difficulties.
This makes it interesting to validate
in silico
homology modeling tools for said
proteins. Correctly modeling the binding site could be the only alternative to obtain
reliable structural models for the drug lead discovery processes. This would yield an
important increase of already available structures in the PDB. It would also further
consolidate homology models in structure-based virtual screening [91-149].
Homology or comparative modeling is the process of generating a structural
model of a target protein from a known template structure. Homologous proteins
evolve from a common ancestor. Due to mutations they can be however
dissimilar. Sequence similarity can be less preserved than structural similarity. In
order to maintain function, structure is typically more conserved than sequence. In
regions close to the surface (loop regions) we observe the most pronounced
dissimilarities. In these regions we can even find changes in the physicochemical
properties of the side chains. There are less frequent changes for residues in the
interior of the proteins. Highly conserved are the center of the proteins (common
core of residues) and the main elements of the secondary structure. Although less
accurate than high-resolution experimental methods, homology modeling can be
helpful in the drug discovery process.
