Three-dimensional pharmacophore modelling studies on antagonists of endothelin receptor ETA

We present 3D pharmacophore modelling and 3D quantitative structure-activity relationship (QSAR) studies on a collection of N-(isoxazolyl) sulphonamide endothelin (ET) receptor antagonists spanning nearly five orders of magnitude in their ET receptor antagonist activity. The alignments needed for the development of 3D QSAR models were obtained using the methods of pharmacophore generation and alignment in the Schrodinger module PHASE. Training and corresponding test sets for the 3D QSAR models were derived by binning the activities into five segments and choosing various percentages randomly within each bin. ETA receptor antagonists training set correlation coefficients (R-2) and mean square errors ranged from 0.7 to 0.97 and 0.2 to 0.6 respectively. The test set predictions were characterized by Q(2) values ranging from 0.3 to 0.55, whereas the Pearson-R factors and root mean square errors varied from 0.6 to 0.75 and 0.2 to 0.8 respectively. Our studies clearly demonstrate that we have derived various models that are predictive for the test sets of ET receptor antagonists. Furthermore, we have employed data scrambling studies of the 3D pharmacophore models of ETA receptor antagonists to validate the QSAR models.