3D-QSAR INVESTIGATION OF THE TRIPOS BENCHMARK STEROIDS AND SOME PROTEIN-TYROSINE KINASE INHIBITORS OF STYRENE TYPE USING THE TDQ APPROACHZ

摘要

The three-dimensional QSAR (TDQ) approach was used to study the Tripos steroid data set and some protein-tyrosine kinase inhibitors of styrene type. The TDQ protocol involves a conformational analysis and rigid object optimization of each investigated structure where the method of partial least squares (PLS) is used as the statistical tool. Compound description is based on either a shape/hydrogen bonding characterization of Compass type or a non-bonded/electrostatic characterization of traditional molecular field analysis (MFA) type. The descriptor points were distributed on a surface or a grid that enclosed the structures. Predictive 3D-QSAR models were derived for both data sets of compounds. However, the predictivities of the derived steroid models of new structures were more limited based on the traditional MFA parametrization compared with the models based on the shape/hydrogen bonding description. For the styrene data set only the TDQ models were predictive compared with the models derived using the traditional MFA protocol involving a fixed, single-conformer, orientation of the structures.