engineering techniques need to keep all the good properties of the
molecules to further identify and remove the unwanted properties.
Because a property of a molecule is often associated directly with its
associated functional groups and chains, parts of the molecules
responsible for the unwanted properties are either removed or
replaced with favorable parts. The process is called “rescaffolding”
or “scaffold hopping” when the central part of the molecule is
replaced with the other chemical motif [15–18]. Scaffold hopping
might be an anticipated outcome of a de novo process in order to
obtain a library of lead molecules with potentially improved
properties.
Identification of new lead molecule using the de novo experi-
ments can be advanced by two approaches: structure-based de novo
computational drug design (SBDND) and ligand-based de novo
drug design (LBDND). SBDND makes use of available tertiary
structure of the target protein, which can be solved experimentally
or computationally. SBDND can be achieved by either a ligand
growing or linking approach. In ligand growing method, a frag-
ment is docked into the binding site of target protein and extended
by adding favorable functional groups to it. The linking method is
similar but instead of docking one, multiple small fragments are
docked within and adjacent to the binding pocket of the target
molecule and linked with each other to form one complex mole-
cule. Availability of the target protein structure helps in screening
fragments at both the search and scoring steps. In the absence of
protein tertiary structure, the structure–activity relationship (SAR)
information about the existing active and inactive compounds can
be exploited using an approach called LBDND. Quantitative struc-
ture–activity relationships based on three-dimensional structures of
molecules (3D QSAR) is an important and widely used method in
LBDND. New Molecules are built using the fragment library con-
structed using known molecules with high biological activity.
This chapter focuses on both concepts of computer based
molecular de novo design methods and general protocols to per-
form the study based on SBDND. By discussing the case studies
involved in successful developments of lead and hit molecules, the
potential of this approach will be highlighted.2 Concepts and Challenges
The basic concept behind identifying a new lead molecule is gen-
erating a new molecule using the fragment of existing active frag-
ments. Therefore, final product depends extensively, first on the
initial material and second on their rearrangement to create some-
thing new with different activity. Any de novo software or algo-
rithm must address three basic questions: assembly or clustering of
candidate fragments, evaluation of their features and activity, andFragment-Based Ligand Designing 125