ABSTRACT
A new 3D descriptor, the local intersection volume (LIV), was developed by our group and applied to the construction of 3D-QSAR models for ligands of the PGI(2) receptor (IP). The target compounds are a set of 42 aromatic heterocyclic derivatives [Meanwell et al., J. Med. Chem. 36 (1993), 3884], which show agonist activities in the IP receptor and are inhibitors of platelet aggregation. The LIV-3D-QSAR models were obtained through the analysis of 30% of the generated conformations for each compound, using a combined Genetic Algorithm (GA) and Partial Least Square (PLS) approach [Rogers and Hopfinger, J. Inf. Comput. Sci. 34 (1994) 854]. Statistically, Model 3 is the best as well as the most comprehensive in a mechanistic sense. Furthermore, it can be applied to design new IP ligands.