PHARMACOPHORE REPRESENTATION AND SEARCHING Pharmacophore Representation and Searching

Pharmacophores can be defined in a number of ways [1, 2]. From a physical standpoint a pharmacophore is a collection of electronic and steric features of a molecule that allow it to exhibit activity against a given biological target. More specifically, a pharmacophore can be defined by a set of atoms and various geometric (distances, angles, torsions) relationships between them. The pharmacophore concept has been used to study a variety of biological targets such as the dopamine D2 receptor [3], HIV integrase [4] and reverse transcriptase [5] and COX2 [6] amongst others. The key component to these studies is the ability to generate databases of 3D conformations of candidate ligands and then search for a given pharmacophore within the database. Currently open source solutions are available for chemical databases [7] and conformer generation. However there are no open source tools for pharmacophore representation and searching. The lack of freely available solutions led to the design of the pharmacophore classes in the CDK. The rest of this article will describe the design of the pharmacophore handling classes and highlight their usage. It should be noted that the current implementation is not a comprehensive pharmacophore modeling solution as it does not provide complex geometric constraints (torsions and excluded volumes) or any graphical or pharmacophore discovery capabilities. However the framework is sufficiently general for these features to be incorporated in the future.