These are some early efforts to explain pharmacophoric patterns that could act as
key features for the design of new chemical entities. Figure 2 shows few early
milestones in the field of emergence of pharmacophore modelling.
Nevertheless, in recent years, many effective pharmacophore modelling
approaches and their contributions to drug discovery have been reported [59]. With
the help of pharmacophoric insights and 3D searching tools, computer-aided drug
design efforts are swiftly gaining efficiency since the 1990s. Still, this approach
encounters many challenges that restrict its success. Pharmacophore approaches
have been widely used in virtual screening, de novo ligand design, lead optimization and multi-target drug design. A range of automated pharmacophore
modelling and screening tools have constantly appeared after the computational
chemistry revolution witnessed in the past couple of decades [60]. Today, pharmacophore screening is one of the apt choices for researchers working in drug
discovery and design.
5 Pharmacophore Model Generation
Pharmacophore models are typically generated either from a group of ligands, by
aligning them and taking out the common interaction features indispensable for
their biological activity. On the other hand, they can be constructed in a
structure-based way, by probing probable interaction points in the receptor binding
pocket, provided the 3D structure of the receptor is reported. The pharmacophore
models can also be generated from a receptor–ligand complex by identifying the
key interactions between the receptor and ligands.
5.1 Ligand-Based Pharmacophore Model Generation
Ligand-based pharmacophore modelling approach is used as a key strategy for
facilitating screening compound databases when there is no three-dimensional
structures are available for the target or receptor, but structure of a set of potent
inhibitors are available. These active molecules are superimposed, and common
pharmacophoric features representing crucial interactions between the ligands and
the common target of these molecules are identified. Firstly, a conformational space
of each of the active ligands is created corresponding to the flexibility of ligands,
followed by their alignment and determination of the important common chemical
features required for the creation of pharmacophore models. Currently, various
automated pharmacophore generators are in use such as Phase [46] (Schrodinger
Inc., http://www.schrodinger.com), HypoGen [61], HipHop [61] (Accelrys Inc.,
http://www.accelrys.com), GASP [62], DISCO [63], GALAHAD [64] (Tripos Inc.,
http://www.tripos.com) and MOE (Chemical Computing Group, http://www.
chemcomp.com) [65]. Several academic programmes [40, 60, 66–68] are also
32
C. Choudhury and G. Narahari Sastry
key features for the design of new chemical entities. Figure 2 shows few early
milestones in the field of emergence of pharmacophore modelling.
Nevertheless, in recent years, many effective pharmacophore modelling
approaches and their contributions to drug discovery have been reported [59]. With
the help of pharmacophoric insights and 3D searching tools, computer-aided drug
design efforts are swiftly gaining efficiency since the 1990s. Still, this approach
encounters many challenges that restrict its success. Pharmacophore approaches
have been widely used in virtual screening, de novo ligand design, lead optimization and multi-target drug design. A range of automated pharmacophore
modelling and screening tools have constantly appeared after the computational
chemistry revolution witnessed in the past couple of decades [60]. Today, pharmacophore screening is one of the apt choices for researchers working in drug
discovery and design.
5 Pharmacophore Model Generation
Pharmacophore models are typically generated either from a group of ligands, by
aligning them and taking out the common interaction features indispensable for
their biological activity. On the other hand, they can be constructed in a
structure-based way, by probing probable interaction points in the receptor binding
pocket, provided the 3D structure of the receptor is reported. The pharmacophore
models can also be generated from a receptor–ligand complex by identifying the
key interactions between the receptor and ligands.
5.1 Ligand-Based Pharmacophore Model Generation
Ligand-based pharmacophore modelling approach is used as a key strategy for
facilitating screening compound databases when there is no three-dimensional
structures are available for the target or receptor, but structure of a set of potent
inhibitors are available. These active molecules are superimposed, and common
pharmacophoric features representing crucial interactions between the ligands and
the common target of these molecules are identified. Firstly, a conformational space
of each of the active ligands is created corresponding to the flexibility of ligands,
followed by their alignment and determination of the important common chemical
features required for the creation of pharmacophore models. Currently, various
automated pharmacophore generators are in use such as Phase [46] (Schrodinger
Inc., http://www.schrodinger.com), HypoGen [61], HipHop [61] (Accelrys Inc.,
http://www.accelrys.com), GASP [62], DISCO [63], GALAHAD [64] (Tripos Inc.,
http://www.tripos.com) and MOE (Chemical Computing Group, http://www.
chemcomp.com) [65]. Several academic programmes [40, 60, 66–68] are also
32
C. Choudhury and G. Narahari Sastry