Molecular Recognition and Drug Design

Prediction of 3D structure of protein or target molecule was quite tedious and uncertain due to many of the protein targets are membrane bound and in case of soluble targets like enzymes, they exist in quite dynamic condition in side the body and its structure-function is tightly regulated by microenvironment. But In recent past this problem was solved at quite extant due to development in techniques like NMR and X-ray diffraction leads to generation of hug databases of protein structures, Along with this development in computational capability have deeply influence the over all process. Now we have number of software and models by which one can predict structure of protein based on just amino acid sequences, classical example if homology modeling of protein folding. In post genomic era high throughput protein expression and structure determination by X-ray diffraction augmented by homology modeling makes key process for new drug development program.
Quantitative Structure Activity Relationship (QSAR) is an area of computational research where virtual model was developed to determines binding proertise of ligand to its target molecules as well as to predict toxicological potential of existing or hypothetical molecules. QSAR is generally employed to establish correlation between structure and electronic properties of ligand molecules which influence its binding to specific drug target and general target molecules. Initially it was used to predict ligand having very high affinity towards target molecules but now it is extended to predict its diffusion, adsorption, toxicity, metabolism and finally elimination. There are different types of QSAR like 1D, 2- (Lill, 2007) (Mller, 2003)D etc based on number of parameters taken for model prediction. Table 1 describes different mode of QSAR and parameters involved in it.
Table:1
(Lill, 2007)
Initially QSAR was developed on single parameters like pKa value or solubility of ligand molecules and based on that prediction was carried out (1D-QSAR). Hansch e’tal has included physic-chemical properties like functional groups and atomic configuration in to it. They also correlated these properties to biological activity of ligand. (2D-QSAR). After 1980s increasing number of 3D structures of proteins makes it more feasible to include three dimensional structure of protein-ligand to understand its interaction. Than after Structure based deign (SBD) becomes routine process for new drug development process. Here after identification of target molecule different ligands were searched and analyzed for its docking to target molecules by process called dynamic optimization (MD). Based on this Technique it is possible to identified best binding mode of any given ligand molecule with target. In 1988 Comparative molecular field analysis (CoMFA) was introduced in QSAR which leads to first time demonstration of structure -function co-relation (3D-QSAR). Development of 3D QSAR made task simpler but later it was realized that 3D QSAR based models are not always gives complete picture but fails to explain processes like Induced fit