QSAR using 2D descriptors and TRIPOS' SIMCA

The combination of 2-dimensional descriptors and classification analysis ha s seen limited use within drug design either due to the general nature of t he descriptors used or by the drive to use only 3D information. We present the use of SIMCA as implemented by TRIPOS in conjunction with our in-house 2D topological descriptors as a means of giving chemically significant anal yses without the need for an alignment step. The TRIPOS method was applied to two published data sets, an in-house data set and two artificial data se ts. The results showed that the structural features deemed to be necessary for the desired activity were identified. These experiments also highlighte d the significant differences between the TRIPOS and literature versions of SIMCA. The potential uses of the SIMCA/2D technique seem limitless as any activity can be categorised.