from fitting into the binding site. That is why inclusion of some excluded volume
features is often recommended which penalizes the molecules’ score if some atoms
or group are placed in positions where they are likely to collide with the active site
atoms.
5.3 Generation of Pharmacophore Models
from the Protein–Ligand Complexes
Protein–ligand complexes produced by X-ray crystallography provide a detailed
picture of the interactions between the ligand and the receptor, showing which
atoms of the ligand are in contact with the receptor along with the atomic coordinates of those atoms. Also, the type of interactions can also be delineated from the
atom types, distances and orientations of the ligand and receptor atoms. The major
interaction that occurs in the receptor–ligand interface is hydrogen bonding. But
other non-covalent interactions such as p–p and cation–p interactions are also
obviously essential for protein–ligand complex formation apart from the hydrogen
bonding. We have extensively looked at the importance of these interactions and
the cooperativity existing among themselves to maintain supramolecular structures
[83–86]. This information is of immense importance to establish a pharmacophore
model from the complex. However, one needs to give attention to the facts that
alternative pharmacophore models are possible within a single binding pocket
owing to the flexibilities of both the active site and the ligands which are capable of
rearranging themselves to accommodate different ligands and also there is a possibility of more than one active sites for a particular receptor. The programmes like
‘LigandScout’ developed by Wolber and Langer [87] and Phase [46] module of
Schrodinger suite generate structure-based pharmacophore models from the protein–ligand complexes given as an input. We will be discussing the steps of generation of pharmacophore models from the protein–ligand complexes by the
LigandScout and Phase, where the former characterizes the pharmacophoric features using kekule’s patterns and the latter prioritizes the features based on the XP
docking energy components.
5.3.1 Pharmacophore Model Generation with LigandScout
With the LigandScout [87] programme, as a first step, the correct molecular
topology of rings and of hybridization state are assigned to the ligands by analysing
the neighbouring atoms followed by assignment of double bonds and Kekule’s
patterns for functional groups such as carboxylic acids and esters, nitro groups,
sulphonyl groups, thio acids, thio acetic esters guanidine-like groups, acetamidine
and phosphinoyl groups functional groups. Next, the pharmacophoric features
based on the hydrogen bonds, electrostatic interactions, charge transfer or
Pharmacophore Modelling and Screening: Concepts, Recent …
37
features is often recommended which penalizes the molecules’ score if some atoms
or group are placed in positions where they are likely to collide with the active site
atoms.
5.3 Generation of Pharmacophore Models
from the Protein–Ligand Complexes
Protein–ligand complexes produced by X-ray crystallography provide a detailed
picture of the interactions between the ligand and the receptor, showing which
atoms of the ligand are in contact with the receptor along with the atomic coordinates of those atoms. Also, the type of interactions can also be delineated from the
atom types, distances and orientations of the ligand and receptor atoms. The major
interaction that occurs in the receptor–ligand interface is hydrogen bonding. But
other non-covalent interactions such as p–p and cation–p interactions are also
obviously essential for protein–ligand complex formation apart from the hydrogen
bonding. We have extensively looked at the importance of these interactions and
the cooperativity existing among themselves to maintain supramolecular structures
[83–86]. This information is of immense importance to establish a pharmacophore
model from the complex. However, one needs to give attention to the facts that
alternative pharmacophore models are possible within a single binding pocket
owing to the flexibilities of both the active site and the ligands which are capable of
rearranging themselves to accommodate different ligands and also there is a possibility of more than one active sites for a particular receptor. The programmes like
‘LigandScout’ developed by Wolber and Langer [87] and Phase [46] module of
Schrodinger suite generate structure-based pharmacophore models from the protein–ligand complexes given as an input. We will be discussing the steps of generation of pharmacophore models from the protein–ligand complexes by the
LigandScout and Phase, where the former characterizes the pharmacophoric features using kekule’s patterns and the latter prioritizes the features based on the XP
docking energy components.
5.3.1 Pharmacophore Model Generation with LigandScout
With the LigandScout [87] programme, as a first step, the correct molecular
topology of rings and of hybridization state are assigned to the ligands by analysing
the neighbouring atoms followed by assignment of double bonds and Kekule’s
patterns for functional groups such as carboxylic acids and esters, nitro groups,
sulphonyl groups, thio acids, thio acetic esters guanidine-like groups, acetamidine
and phosphinoyl groups functional groups. Next, the pharmacophoric features
based on the hydrogen bonds, electrostatic interactions, charge transfer or
Pharmacophore Modelling and Screening: Concepts, Recent …
37