Chemistry Reference
In-Depth Information
There are many available methods for
in silico
bioisosteric replacements and
scaffold hopping to identify groups with similar properties. These methods apply
various techniques of chemoinformatics as the characterization of groups by a
large number of calculated descriptors and identification of bioisosteric pairs
based on similarity of properties. Langdon
et al
. [3] report different types of
descriptors that have been described and compiling distinct molecular properties
for different applications. These descriptors can be grouped into three main
categories: topological descriptors (2D descriptors that retain information on
molecular characteristics such as atoms, types of atoms, substructures, and how
these characteristics are connected with the molecules), topographic descriptors
(also descriptors of molecular characteristics, but retain information from
geometric of these features and are, thus, 3D descriptors) and surface-based or
field-based descriptors (are also 3D descriptors, but encode information on the
surface of the molecule or molecular field).
Bergmann and coworkers [29] presented a GRID-based method for scaffold
hopping, the SHOP methodology. This similarity search approach based on GRID
identifies new scaffolds in a database by analyzing the similarity of their three-
dimensional structures with those query scaffolds. The aim of such approach is to
find substituents for consulting the scaffold while the geometry, form and patterns
of interaction of the central fragment of the query ligand are retained. For this, the
descriptors used for these searches are specific anchor points, where the anchor
refers to positions on the scaffolds. In the SHOP software, scaffolds are described
geometrically in accordance with the distance and dihedral angles between their
anchor points. The scaffold's shape is based on analysis of these distances
between each anchor point and the surface of the scaffold calculated with MIFs.
The GRID program is then used to obtain the description of the scaffold's shape
by calculating the energies of the interaction in a grid-shaped box around the
scaffold, thus, obtaining the distances between each anchor point and each grid
point with an energy value that is registered. The most favorable energy values are
selected. These anchor points selected are then used in the search step within
molecular databases by similarity. Thus, with the scaffolds selected within the
database it is possible to create new scaffolds which are able to retain the desired
ligand properties of the original bioactive scaffold.
