Generation of pharmacophores and classification of drugs using protein-ligand complexes

Pharmacophore identification is a very
important step in de novo design, lead
optimization, chemogenomics, and virtual
screening of drugs. Unfortunately,
the high cost of comercial software for
pharmacophore detection is a common
limiting factor for researchers with limited
funding. This paper presents a set of
freely available perl routines that were
designed to help in the process of 3D
pharmacophore identification and QSAR
studies. These routines also allowed the
classification of ligands based on their
tridimensional similarity and binding
mechanism. The family of phosphodiesterases
and their inhibitors were used as
test model.