ABSTRACT
Pharmacophoric mapping is of great value in generating new chemical lead structures, especially when a limited number of compounds are available or when different chemical classes are used (see Chapter 4). A structural fit, reflected by the 3-D geometry of a structure in its active conformation is a necessary but not sufficient cause of activity, since electronic and hydrophobic forces between ligand and receptor are required for the response. An inherent limitation with pharmacophoric modeling techniques like the active analog approach is their inability to quantitatively describe the biological effect, i.e. one can usually only distinguish active from inactive compounds. In the process of optimizing a lead structure, it is necessary to utilize the information from quantitative activity data and from other structural properties in a more efficient way in order to predict more active congeners. Furthermore, quantitative structure-activity relationships (QSAR) can provide a great deal of information regarding the nature of ligand-target protein interactions. In series of homologous derivatives, various quantitative structureactivity analyses utilizing linear free energy relationships, multiple linear regression, and pattern recognition techniques have been applied. Furthermore, recent progress has been made in combining molecular modeling and statistical models, which allows for handling of non-congeneric series.
