The usefulness, in pharmacological classification, of complementary pattern-recognition techniques and structure modelling as afforded by the iterative collation of multiple-trial data in data banks.
In this emerging information age no significant limits can be envisaged to the immense resources of knowledge that pharmacological sciences can draw upon by systematically applying multivariate pattern-recognition techniques to those data banks which can be organized internationally with better standardization of descriptors, i.e. by parametrizing observations and evaluating monitoring in experimental, biological assays, clinical trials and postmarketing surveillance. Even conventionally or habitually adopted references and communalities such as traditional drug profiles and receptor models may be iteratively re-checked and suitably adapted so as to take account of more adequate, up-to-date analytical techniques, fresh biological ideas and new advances in terms of physiological refinements. An attempt may also be made to modify the chemicophysical relationships and patterns currently traced on the basis of what are held to be quantitative structure-activity oversimplifications. The present paper focuses upon specifying a number of standardization criteria in conventional assays and upon submitting multiple biological features to monitoring; in addition, it gives some picture of the new trends the approach can offer and draws attention to the more relevant, innovative literature references.