| Two structure-based drug design virtual screening approaches have been designed in this study to identify lead candidates during drug discovery. The first approach is the "Design of a docking-based virtual high throughput screening system", which allows identification of target specific lead molecules and library prioritization. The idea is to dock multiple ligands in any number of desired proteins simultaneously using a user-friendly docking web application for conducting a comparative study to identify target-specific ligands. This user-friendly CGI/Perl application allows any number of medicinal or synthetic chemists to remotely use a computational tool traditionally reserved for experts in molecular modeling, thereby increasing the efficiency of drug design process. The docking scores are hyperlinked to a graphical output, which highlights the 3D structure of the protein region involved in ligand binding, to visualize whether the ligand is docking in the desired mode. The second approach uses the novel technique "CATveat" for performing scaffold hopping to identify new drug candidate molecules. The basic principle of CATveat is to preserve certain desirable functional groups responsible for the favorable interactions of the drug with the protein, similar to existing structure-based software 'Caveat'. However, the CATveat approach uses a series of programs, including FrameMaker and Catalyst, which aid in overcoming the shortcomings associated with Caveat. The application of FrameMaker to the ligand database reduces the torsional degree of freedom of ligand molecules by collapsing the acyclic side-chains to synthetic handles, marked by fluorine atoms bonded to the core ring framework (scaffold). This processing of database, besides helping the scaffold to cover an elaborate chemical space during multi-conformer database generation, improves the synthesizability of the hits obtained by CATveat. The use of Catalyst facilitates the generation of a multi-conformer database, development of a shape-based query and explicit mapping of hydrogen bonds as bond vectors, features which are not found in Caveat. The validity and usefulness of CATveat are demonstrated by applying it to obtain novel scaffolds for design of dihydrofolate reductase (DHFR) enzyme inhibitors for Pneumocystis carinii, a fungus responsible for causing crippling pneumonia to immunocompromised AIDS patients. |
